JPWO2018235938A1 - Methods for sequencing and analyzing nucleic acids - Google Patents

Abstract

The present invention provides a method for correcting errors that occur in a digital nucleic acid quantification method using a molecular barcode. Specifically, a method of identifying the mispairing of the index sequence and the molecular barcode according to the detection frequency, a method of identifying the molecular barcode having base substitutions classified into the same cluster, and a fixed base and a random base. Methods of identifying molecular barcodes with insertions or deletions using molecular barcodes containing are provided.

Description

Reference to related applications

  This application is an application that enjoys the benefit of the priority of US Provisional Application No. 62/523857 (filing date: June 23, 2017), and the entire provisional application is incorporated herein by reference. I shall.

  The present invention relates to methods for sequencing and analyzing nucleic acids.

  With the development of next-generation sequencer platform, it has become possible to simultaneously analyze the sequences of a large number of nucleic acid types in a single run. By adding a unique molecular barcode to each nucleic acid molecule present in a sample, the number of unique molecular barcode types can be made to correspond to the number of nucleic acid molecules. Has opened up the way for digital quantification (Patent Document 1 and Non-Patent Document 1). By using a molecular barcode as a random base and lengthening the nucleotide sequence, it becomes possible to easily add a great diversity to the barcode sequence, and the dynamic range of nucleic acid molecules that can be digitally quantified is expanded (Patent Document 1 and Non-Patent Document 1).

  However, in digital quantification, the sequence of the molecular barcode may change during the analysis, and the newly generated molecular barcode may affect the quantification accuracy of nucleic acid molecules. However, if the sequence of the molecular barcode is randomly designed, it is difficult to understand that the sequence has changed. In addition, it was difficult to analyze what kind of error could occur in digital quantification due to the random arrangement of the molecular barcode, and it was also difficult to provide a solution.

U.S. Patent No. 9,260,753

Shiroguchi, K. et al., Proc. Natl. Acad. Sci. USA 109, 1347-1352 (2012).

  The present invention provides methods for sequencing and analyzing nucleic acids.

In the digital quantification method of a target nucleic acid molecule using an index and a barcode, the present inventors have found that when a plurality of samples are mixed to quantify the target nucleic acid molecule, the index is a nucleic acid derived from a different sample than expected. It has been clarified that a misindex, which is added to, may occur. The inventors have also determined that if two different indexes are added to the same barcode, the most frequent pair is regarded as the correct pair and any or all other pairs are excluded as a miss index. , Clarified that the accuracy of the digital quantitative method can be improved.
The present inventors have found that when counting the number of types of barcode sequences, mutations (eg, insertions, substitutions, and deletions) occur in the barcode sequences, and sequences that should be judged to be the same are identified as different sequences. It became clear that perceived problems could occur. The present inventors have clarified that the accuracy of the digital quantification method can be improved by clustering sequences having a certain sequence similarity into a group and quantifying the target nucleic acid molecule based on the number of clusters.
The present inventors have revealed that when digitally counting nucleic acids, a problem of misidentifying a template may occur. The inventors have also determined that when two different nucleic acid sequences of interest are added to the same barcode, the most frequent pair is the correct pair and any or all other pairs are excluded as misidentifications. By doing so, it was clarified that the accuracy of the digital quantitative method can be improved.

That is, according to the present invention, the following inventions are provided.
(1A) A method for analyzing a nucleic acid, which comprises:
(I) subjecting a mixture of a plurality of target nucleic acid molecules having a molecular barcode and an index to sequencing to obtain sequence information;
(II) A sequence having a specific index or a sequence similar thereto and / or a sequence having a specific molecular barcode or a sequence similar thereto is selected from the sequence information obtained in (I) above, and selected. Creating a group with the arranged array,
(III) in the group created in (II) above, determining the pair of the index and the molecular barcode with the highest detection frequency as the correct pair of the index and the molecular barcode,
Including the method.
(2A) The method according to (1A) above, wherein the nucleic acid molecule of interest to which at least the molecular barcode is added is subjected to amplification before step (I).
(3A) A sequence similar to the sequence having the specific molecular barcode in the step (II) is a sequence containing a mismatch base having a predetermined number of bases or less in the molecular barcode sequence portion with the sequence having the specific molecular barcode. The method according to (1A) or (2A) above.
(4A) The method according to any one of (1A) to (3A) above, wherein the molecular barcode has a fixed base at a specific position.
(5A) A sequence similar to the sequence having the specific molecular barcode in step (II) contains the fixed base at the specific position, and / or the position of the fixed base is shifted from the specific position. The method according to (4A) above, which is selected based on
(6A) The method according to (4A) above, further comprising excluding a sequence having a molecular barcode that does not include the fixed base at the specific position from analysis.
(7A) In step (III), a pair of an index and a molecular barcode other than the determined correct pair is determined and excluded as a mispair of the index and the molecular barcode.
The method according to any one of (1A) to (5A) above.
(8A) The method further comprises the step of determining the number of the target nucleic acid molecule contained in the sample from which the target nucleic acid molecule is derived, based on the number of groups created by the sequence having a specific molecular barcode or a sequence similar thereto. , The method according to any one of (1A) to (7A) above.
(9A) A method for analyzing nucleic acid, which comprises:
(I) subjecting a mixture of a plurality of nucleic acid molecules having a molecular barcode to sequencing to obtain sequence information;
(II) a step of selecting a sequence having a specific molecular barcode or a sequence similar thereto from the sequence information obtained in (I) above, and creating a group by the selected sequence;
Including the method.
(10A) A sequence similar to the sequence having the specific molecular barcode in the step (II) is a sequence containing a mismatch base having a predetermined number of bases or less with the sequence having the specific molecular barcode in the molecular barcode sequence portion. The method according to (9A) above.
(11A) The method according to (9A) or (10A) above, wherein the molecular barcode has a fixed base at a specific position.
(12A) A sequence similar to the sequence having the specific molecular barcode in step (II) contains the fixed base at the specific position, and / or the position of the fixed base is shifted from the specific position. The method according to (11A) above, wherein the method is selected based on:
(13A) The method according to (11A) above, further including a step of excluding a sequence having a molecular barcode that does not include the fixed base at the specific position from analysis.
(14A) further comprising the step of determining the number of the target nucleic acid molecule contained in the sample from which the target nucleic acid molecule is derived, based on the number of groups created by the sequence having a specific molecular barcode or a sequence similar thereto. The method according to any one of (9A) to (13A) above.
(15A) The method according to any one of (9A) to (14A) above, wherein the nucleic acid molecule of interest to which at least the molecular barcode is added is subjected to amplification before step (I).
(16A) A method for analyzing a nucleic acid, which comprises:
(I) subjecting a mixture of a plurality of nucleic acid molecules having a molecular barcode having a fixed base at a specific position to sequencing to obtain sequence information;
(IIa) excluding from the analysis a sequence having a molecular barcode that does not include the fixed base at the specific position;
(IIb) In step (I) or after step (I), a step of obtaining sequence information consisting of a sequence containing the fixed base at the specific position; or (IIc) the step (II) above (I). ) Further comprising the step of selecting a sequence having a specific molecular barcode or a sequence similar thereto from the sequence information obtained in step (4), and forming a group with the selected sequence, and in the step (II) or the step (II) After II), obtaining a group of sequences comprising the fixed base at the particular position;
Including the method.

According to the present invention, the following inventions are also provided.
(1B) Obtained by sequencing using a mixture of a plurality of samples containing a target nucleic acid molecule having a unique index for each sample containing a plurality of nucleic acid molecules and a unique or arbitrary molecular barcode for each nucleic acid molecule A method for determining the correct pair or mispair of the index and the molecular barcode added to the target nucleic acid molecule from the sequence information,
(E) From the obtained sequence information, a sequence having a specific index or a sequence similar thereto, a sequence having a specific molecular barcode or a sequence similar thereto, or a sequence containing a target nucleic acid molecule or a sequence similar thereto Selecting sequences and creating groups with the selected sequences;
(F) In the group created in (E) above, the pair of the index and the molecular barcode having the highest detection frequency is determined as the correct pair of the index and the molecular barcode, and / or the index and the molecule having the low detection frequency are determined. Determining at least any one or all of the barcode pairs as an index and a molecular barcode mispair.
Including the method.
(2B) In step (E), an array having a specific index is selected to create a group for each index,
In the step (F), regarding a nucleic acid sequence having a molecular barcode that appears in a plurality of groups, the pair of barcode and index having the largest number of reads is determined as the correct pair of barcode and index, or the detection frequency Determine the highest index and molecular barcode pair as the correct index and molecular barcode pair,
The method according to (1B) above.
(3B) In step (E), a sequence having a specific molecular barcode is selected to create a group for each molecular barcode,
In the step (F), the pair of the index and the molecular barcode having the highest detection frequency among the created groups is determined as the correct pair of the index and the molecular barcode.
The method according to (1B) above.
(4B) In step (E), a sequence containing the sequence of the target nucleic acid molecule is selected to form a group,
In the step (F), a sequence having a specific index is further selected from the group to create a subgroup, and a barcode having the largest number of reads is selected for a nucleic acid sequence having a molecular barcode that appears in a plurality of subgroups. Determining the index pair as the correct barcode and index pair, or determining the most frequently detected index and molecular barcode pair as the correct index and molecular barcode pair,
The method according to (1B) above.
(5B) In step (E), a sequence containing the sequence of the target nucleic acid molecule is selected to form a group,
In the step (F), a molecule having a specific molecular barcode is further selected from the group to create a subgroup, and the index with the highest detection frequency and the molecular barcode pair is indexed in one created subgroup. And determine the correct pair of molecular barcodes,
The method according to (1B) above.
(6B) In the step (F), at least any one or all of the pairs of the index and the molecular barcode other than the determined correct pair are determined as the mispair of the index and the molecular barcode.
The method according to (2B) to (5B) above.
(7B) In step (E), a molecule having a specific index is selected to create a group for each index,
In the step (F), with respect to the sequences having the molecular barcodes appearing in a plurality of groups, at least one or all of the indices and the molecular barcode pairs having a low detection frequency are determined as the mispairs of the molecular barcodes. ,
The method according to (1B) above.
(8B) In step (E), a sequence having a specific molecular barcode is selected to create a group for each molecular barcode,
Determining a pair of an index and a molecular barcode having a low detection frequency among the group created in the step (F) as at least one or all of the indexes and the molecular barcode.
The method according to (1B) above.
(9B) In step (E), a sequence containing the target nucleic acid molecule is selected to form a group,
In step (F), a molecule having a specific index is further selected from the group to form a subgroup, and a nucleic acid molecule having a molecular barcode that appears in a plurality of subgroups has a low detection frequency and a molecular barcode. Determine at least any one or all of the pairs of as the pair of index and molecular barcode,
The method according to (1B) above.
(10B) In step (E), molecules containing the target nucleic acid molecule are selected to form a group,
In the step (F), a molecule having a specific molecular barcode is further selected from the group to create a subgroup, and at least one of a pair of an index and a molecular barcode having a low detection frequency in one created subgroup. Determining one or all as the index and molecular barcode mispair,
The method according to (1B) above.
(11B) In step (E), the step of creating a group creates a group by clustering molecules that are presumed to have the same sequence judged based on the sequence identity or similarity as a group. Done by
The method according to (1B) to (10B) above.
(12B) In step (E), the clustering is
(I) In the sequence of the molecular barcode portion, a nucleic acid molecule group having the same sequence as the unique molecular barcode sequence is classified into the same cluster;
(Ii) In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence having a mismatch of up to 1 base with the sequence of the unique molecular barcode is classified into the same cluster;
(Iii) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a unique sequence of the molecular barcode and a sequence having a mismatch of up to 2 bases is classified into the same cluster; or (iv) the molecular barcode In a partial sequence, a group of nucleic acid molecules having a unique molecular barcode sequence and a sequence having a mismatch of up to 3 bases is classified into the same cluster.
The method according to (11B) above.
(13B) In step (E), the clustering is
In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence sequenced as having a base insertion or deletion (indel) is classified into the same cluster,
The method according to (11B) or (12B) above.
(14B) In the step (E), the clustering is
Performed on a group of nucleic acid molecules obtained by excluding the sequence sequenced as having a base insertion or deletion (indel) in the sequence of the molecular barcode portion,
The method according to (11B) or (12B) above.
(15B) The insertion or deletion of the above-mentioned bases in the positions of each of one or more fixed bases arranged in all the molecular barcode sequences linked to the nucleic acid molecule, and in the sequence of the sequenced molecular barcode sequence portion. The method according to (13B) or (14B) above, further comprising identifying by the difference in position from the position of each of the one or more fixed bases.
(16B) Obtained by sequencing using a mixture of a plurality of samples containing a target nucleic acid molecule having a unique index for each sample containing a plurality of nucleic acid molecules and a unique or arbitrary molecular barcode for each nucleic acid molecule A method for determining the number of target nucleic acid molecules contained in a specific original sample from the sequence information
(E) selecting a nucleic acid molecule containing the sequence of the target nucleic acid molecule from the obtained sequence information;
(F) clustering the nucleic acid molecules selected in (e) above for each unique molecular barcode sequence, and then identifying clusters having a plurality of sequences in the index nucleic acid molecule portion;
(G) In each of the clusters identified in (f) above, the pair of the index and the molecular barcode with the highest detection frequency is identified as the correctly indexed target nucleic acid molecule, and the other pairs of the index and the molecular barcode are identified. Deciding to be a mispair,
Including,
The number of unique molecular barcode sequence types linked to the correctly indexed target nucleic acid molecule (or the number of correctly indexed clusters of the target nucleic acid molecule) is contained in the sample corresponding to the index. Is the number of
Method.
(17B) In (f) above, the clustering is
(I) In the sequence of the molecular barcode portion, a nucleic acid molecule group having the same sequence as the unique molecular barcode sequence is classified into the same cluster;
(Ii) In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence having a mismatch of up to 1 base with the sequence of the unique molecular barcode is classified into the same cluster;
(Iii) by classifying into the same cluster nucleic acid molecule groups having a sequence having a mismatch of up to 2 bases with the sequence of the unique molecular barcode in the sequence of the molecular barcode portion; or (iv) the molecular barcode In a partial sequence, a group of nucleic acid molecules having a unique molecular barcode sequence and a sequence having a mismatch of up to 3 bases is classified into the same cluster.
The method according to (16B) above.
(18B) In (e) above, the clustering is
In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence sequenced as having a base insertion or deletion (indel) is classified into the same cluster,
The method according to (16B) or (17B) above.
(19B) In (e) above, the clustering is
Performed on a group of nucleic acid molecules obtained by excluding the sequence sequenced as having a base insertion or deletion (indel) in the sequence of the molecular barcode portion,
The method according to (16B) or (17B) above.
(20B) The insertion or deletion of the above-mentioned bases in the position of each of one or more fixed bases arranged in all the molecular barcode sequences linked to the nucleic acid molecule and in the sequence of the sequenced molecular barcode sequence portion. The method according to (18B) or (19B) above, which further comprises specifying by the difference from the position of each of the one or more fixed bases.
(21B) In a digital quantification method of a target nucleic acid molecule using a barcode sequence, based on the information of the obtained nucleic acid sequence, the sequence possessed by the molecular barcode after mutation is grouped together with other sequences having sequence similarity. A method of estimating the number of target nucleic acid molecules based on the obtained number of clusters.
(22B) The method according to (21B) above, wherein the clustering is
(I) In the sequence of the molecular barcode portion, a nucleic acid molecule group having the same sequence as the unique molecular barcode sequence is classified into the same cluster;
(Ii) In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence having a mismatch of up to 1 base with the sequence of the unique molecular barcode is classified into the same cluster;
(Iii) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a unique sequence of the molecular barcode and a sequence having a mismatch of up to 2 bases is classified into the same cluster; or (iv) the molecular barcode The method is performed by grouping nucleic acid molecules having a sequence having a mismatch of up to 3 bases with a sequence of a unique molecular barcode in a partial sequence into the same cluster.
(23B) The method according to (21B) or (22B) above, wherein the clustering comprises inserting bases (for example, up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of the molecular barcode portion, or A method, which is performed by grouping nucleic acid molecules having sequences sequenced as having an indel into the same cluster.
(24B) The method according to any one of (21B) to (23B) above,
The position of each of one or more fixed bases arranged in all the molecular barcode sequences linked to the nucleic acid molecule and the position of each of the one or more fixed bases in the sequence of the sequenced molecular barcode sequence portion The insertion or deletion (indel) of the base by identifying by comparing
Clustering is performed by classifying into the same cluster nucleic acid molecules having a sequence sequenced as having a base insertion or deletion (indel) in the sequence of the molecular barcode portion, or
A method in which clustering is performed on a group of nucleic acid molecules obtained by excluding a sequence sequenced as having a base insertion or deletion (indel) in a sequence of a molecular barcode portion.
(25B) A method for detecting insertion or deletion (indel) of a base in a barcode in a digital quantification method for a nucleic acid molecule of interest using a barcode sequence, wherein all molecular barcode sequences linked to the nucleic acid molecule By identifying the relative position of each of the one or more fixed bases located within and the position of each of the one or more fixed bases in the sequence of the sequenced molecular barcode sequence portion. A method comprising detecting a base insertion or indel.

According to the present invention, the following inventions are also provided.
(1C) An index unique to each sample containing a plurality of nucleic acid molecules (meaning an index sequence nucleic acid molecule, and may include a plurality of types of index nucleic acid molecules if it is unique to each sample) and unique to each nucleic acid molecule Obtained by sequencing (ie, multiplex sequencing) with a mixture of multiple samples containing a nucleic acid molecule of interest (eg, DNA or RNA) to which any or any molecular barcode (barcode sequence nucleic acid molecule) has been added From the sequence information provided, a method for determining the correct pair or mispair of the index and the molecular barcode added to the target nucleic acid molecule,
(A) separately obtaining a plurality of samples containing nucleic acid molecules (for example, DNA or RNA), and {at least one of the samples contains a target nucleic acid molecule},
(B) {for example, in each of the obtained plurality of samples}, before amplifying the nucleic acid molecule contained in the sample, each nucleic acid molecule of interest is linked with a unique or arbitrary molecular barcode of each nucleic acid molecule, Obtaining a target nucleic acid molecule in which different molecular barcodes are linked,
(C) {For example, before mixing a plurality of samples} A target nucleic acid molecule in which a unique index is added to a target nucleic acid molecule for each sample containing a plurality of target nucleic acid molecules and a different index is linked to each derived sample And (step C may be performed after step B, or step B may be performed after step C; and the nucleic acid molecule may be amplified after step B or C to amplify the nucleic acid molecule of interest. Amplification product) can be obtained),
(D) In a mixture containing the amplification products of the nucleic acid molecules obtained after (B) and (C) above (the sample is mixed after the step (C), and after the sample is mixed, the step (B ) May be carried out, and all the samples may be mixed after carrying out step (B), and the amplification product of the nucleic acid molecule linked with the molecular barcode is obtained after step (B), The sample may be mixed before obtaining the amplification product, and the sample containing the amplification product may be mixed after obtaining the amplification product), the index unique to each sample and the unique or arbitrary to each nucleic acid molecule of interest. Sequence of the nucleic acid molecule to which the molecular barcode is added to determine the sequence of the index portion, the sequence of the molecular barcode portion, and the sequence of the target nucleic acid molecule portion linked thereto, if necessary, for each nucleic acid molecule. When,
(E) From the obtained sequence information {for example, it can be performed based on sequence identity or similarity}, a sequence having a specific index or a sequence similar thereto, a sequence having a specific molecular barcode, or Selecting a sequence similar to this, or a sequence containing the nucleic acid molecule of interest or a sequence similar thereto, and creating a group with the selected sequence,
(F) In the group created in (E) above, the pair of the index and the molecular barcode having the highest detection frequency is determined as the correct pair of the index and the molecular barcode, and / or the index and the molecule having the low detection frequency are determined. A pair of barcodes (for example, a pair having a detection frequency lower than a certain reference value, and the certain reference value is 99.5% or less, 99% or less, 90% or less, 80% or less, 70% or less, 60% or less in a group. Examples include, but are not limited to, values of% or less, 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less. At least any one or all of), and determining that at least one or all of them are mispairs of the index and the molecular barcode,
Including the method.
(2C) In step (E), an array having a specific index is selected to create a group for each index,
In the step (F), regarding a nucleic acid sequence having a molecular barcode that appears in a plurality of groups, the pair of barcode and index having the largest number of reads is determined as the correct pair of barcode and index, or the detection frequency Determine the highest index and molecular barcode pair as the correct index and molecular barcode pair,
The method according to (1C) above.
(3C) In step (E), a sequence having a specific molecular barcode is selected to create a group for each molecular barcode,
In the step (F), the pair of the index and the molecular barcode having the highest detection frequency among the created groups is determined as the correct pair of the index and the molecular barcode.
The method according to (1C) above.
(4C) In step (E), a sequence containing the sequence of the target nucleic acid molecule is selected to form a group,
In the step (F), a sequence having a specific index is further selected from the group to create a subgroup, and a barcode having the largest number of reads is selected for a nucleic acid sequence having a molecular barcode that appears in a plurality of subgroups. Determining the index pair as the correct barcode and index pair, or determining the most frequently detected index and molecular barcode pair as the correct index and molecular barcode pair,
The method according to (1C) above.
(5C) In step (E), a sequence containing the sequence of the target nucleic acid molecule is selected to form a group,
In the step (F), a molecule having a specific molecular barcode is further selected from the group to create a subgroup, and the index with the highest detection frequency and the molecular barcode pair is indexed in one created subgroup. And determine the correct pair of molecular barcodes,
The method according to (1C) above.
(6C) In step (F), at least any one or all of the pairs of the index and the molecular barcode other than the determined correct pair are determined as the pair of the index and the molecular barcode.
The method according to any one of (2C) to (5C) above.
(7C) In step (E), a molecule having a specific index is selected to create a group for each index,
In the step (F), with respect to the sequences having the molecular barcodes appearing in a plurality of groups, a pair of an index and a molecular barcode having a low detection frequency (for example, a pair having a detection frequency lower than a certain reference value and having a constant The reference value includes, but is not limited to, values of 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less in the group. Of at least any one or all of the indices and the mispair of the molecular barcode.
The method according to (1C) above.
(8C) In step (E), a sequence having a specific molecular barcode is selected to create a group for each molecular barcode,
A pair of an index and a molecular barcode having a low detection frequency in the group created in the step (F) (for example, a pair having a detection frequency lower than a certain reference value, and the certain reference value is 50% or less in the group). , 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less, but are not limited to these values. Good)) as at least one and / or all mispairs of the index and the molecular barcode,
The method according to (1C) above.
(9C) In step (E), a sequence containing the target nucleic acid molecule is selected to form a group,
In step (F), a molecule having a specific index is further selected from the group to form a subgroup, and a nucleic acid molecule having a molecular barcode that appears in a plurality of subgroups has a low detection frequency and a molecular barcode. Pair (for example, a pair with a detection frequency lower than a certain reference value, and a certain reference value is 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less in a group. , A value of 1% or less, but not limited to these, and may be, for example, a pair of detection frequencies after the second. decide,
The method according to (1C) above.
(10C) In step (E), molecules containing the target nucleic acid molecule are selected to form a group,
In the step (F), a molecule having a specific molecular barcode is further selected from the group to create a subgroup, and a pair of an index and a molecular barcode having a low detection frequency (for example, a fixed value) is created in one created subgroup. A pair of detection frequency lower than the reference value of, a certain reference value is 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less in the group However, the present invention is not limited to these, and may be, for example, at least any one or all of the second and subsequent detection frequency pairs.
The method according to (1C) above.
(11C) In step (E), the step of creating a group had an identical sequence {preferably in the sequence of the molecular barcode portion} having an identical sequence determined on the basis of sequence identity or similarity {eg, The sequence may be changed by any one of the steps (A) to (D)} is performed by clustering the molecules that are presumed to form a group,
The method according to (1C) to (10C) above.
(12C) In the step (E), the clustering is
(I) In the sequence of the molecular barcode portion, a nucleic acid molecule group having the same sequence as the unique molecular barcode sequence {ie, Distance = 0} is classified into the same cluster;
(Ii) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a sequence having a mismatch of up to 1 base with the sequence of the unique molecular barcode (ie, Distance = 1) is classified into the same cluster;
(Iii) In the sequence of the molecular barcode portion, the nucleic acid molecule group {ie, Distance = 2} having a sequence having a mismatch of up to 2 bases with the sequence of the unique molecular barcode is classified into the same cluster; Or (iv) In the sequence of the molecular barcode part, the nucleic acid molecule group {ie, Distance = 3} having a sequence having a mismatch of up to 3 bases with the sequence of the unique molecular barcode is classified into the same cluster. ,
The method according to (11C) above.
(13C) In the step (E), the clustering is
Grouping nucleic acid molecules having sequences sequenced in the sequence of the molecular barcode portion as having insertions or deletions (indels) of bases (eg up to 1 base, up to 2 bases, or up to 3 bases) into the same cluster Is done by
The method according to (11C) or (12C) above.
(14C) In the step (E), the clustering is
Nucleic acid molecule group obtained by excluding the sequence sequenced as having insertion or deletion (indel) of bases (eg, up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of the molecular barcode portion Done against,
The method according to (11C) or (12C) above.
(15C) The insertion or deletion of the base is one or more (eg, 1, 2, 3, 4, 5, or 6) arranged in all molecular barcode sequences linked to the nucleic acid molecule. One or more) fixed bases and the position of each of the one or more fixed bases in the sequence of the sequenced molecular barcode sequence portion is further specified by the position difference. {Eg, each fixed base can be designed to be any one base selected from the group consisting of A, T, G and C; or a combination of A and T, A and T G combination, A and C combination, T and G combination, T and C combination, G and C combination, A and T and G combination, A and T and C combination, A and G and Consists of a combination with C and a combination of T, G and C It may be designed to be a base selected at random from bases contained in any one combination selected from}.
(16C) A target nucleic acid molecule having a unique index (index sequence nucleic acid molecule) for each sample containing a plurality of nucleic acid molecules and a unique or arbitrary molecular barcode (barcode sequence nucleic acid molecule) added to each nucleic acid molecule (for example, , DNA or RNA) is used to determine the number of target nucleic acid molecules contained in a particular original sample from the sequence information obtained by sequencing (ie, multiplex sequencing) using a mixture of a plurality of samples. How to decide,
(A) separately obtaining a plurality of samples containing nucleic acid molecules (for example, DNA or RNA); {at least one of the samples contains a target nucleic acid molecule},
(B) A target nucleic acid molecule in which, before amplification of the nucleic acid molecule contained in the sample, an arbitrary molecular barcode is linked to each target nucleic acid molecule in each of the obtained plurality of samples, and different molecular barcodes are linked to each To obtain
(C) Before mixing a plurality of samples, a unique index for each sample containing a plurality of target nucleic acid molecules is added to the target nucleic acid molecule to obtain a library of target nucleic acid molecules in which different indexes are linked to each derived sample. Step (the order of Step B and Step C may be either first; or, after Step B or C, the nucleic acid molecule can be amplified to obtain an amplification product of the target nucleic acid molecule),
(D) In a mixture containing amplification products of the nucleic acid molecules obtained after (B) and (C) above (the sample is mixed after the step (C), and after the sample is mixed, the step (B ) May be carried out, and all the samples may be mixed after carrying out step (B), and the amplification product of the nucleic acid molecule linked with the molecular barcode is obtained after step (B), The sample may be mixed before obtaining the amplification product, and the sample containing the amplification product may be mixed after obtaining the amplification product), the index unique to each sample and the unique or arbitrary to each nucleic acid molecule of interest. Sequencing the nucleic acid molecule to which the molecular barcode is added to identify the sequence of the index portion, the sequence of the molecular barcode portion and the sequence of the nucleic acid molecule portion linked thereto for each nucleic acid molecule,
(E) selecting a nucleic acid molecule containing the sequence of the target nucleic acid molecule from the obtained sequence information;
(F) clustering the nucleic acid molecules selected in (e) above for each unique molecular barcode sequence, and then identifying clusters having a plurality of sequences in the index nucleic acid molecule portion;
(G) In each of the clusters identified in (f) above, the pair of the index and the molecular barcode with the highest detection frequency is identified as the correctly indexed target nucleic acid molecule, and the other pairs of the index and the molecular barcode are identified. Deciding to be a mispair,
{May further include determining that the index is incorrect in Mispair},
The number of unique molecular barcode sequence types linked to the correctly indexed target nucleic acid molecule (or the number of correctly indexed clusters of the target nucleic acid molecule) is contained in the sample corresponding to the index. Is the number of
Method.
(17C) In (f) above, the clustering is
(I) In the sequence of the molecular barcode portion, a nucleic acid molecule group having the same sequence as the unique molecular barcode sequence is classified into the same cluster;
(Ii) In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence having a mismatch of up to 1 base with the sequence of the unique molecular barcode is classified into the same cluster;
(Iii) by classifying into the same cluster nucleic acid molecule groups having a sequence having a mismatch of up to 2 bases with the sequence of the unique molecular barcode in the sequence of the molecular barcode portion; or (iv) the molecular barcode In a partial sequence, a group of nucleic acid molecules having a unique molecular barcode sequence and a sequence having a mismatch of up to 3 bases is classified into the same cluster.
The method according to (16C) above.
(18C) In (e) above, the clustering is
Nucleic acid molecules having sequences sequenced in the sequence of the molecular barcode portion as having insertions or deletions (indels) of bases (eg, up to 1 base, up to 2 bases, or up to 3 bases) into the same cluster. Done by classifying,
The method according to (16C) or (17C) above.
(19C) In (e) above, the clustering is
A nucleic acid molecule obtained by excluding a sequence sequenced as having an insertion or deletion (indel) of bases (for example, up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of a molecular barcode portion. Performed on a group,
The method according to (16C) or (17C) above.
(20C) The insertion or deletion of the base is one or more (eg, 1, 2, 3, 4, 5, or 6) arranged in all molecular barcode sequences linked to the nucleic acid molecule. 20 or more) and the position of each of the one or more fixed bases in the sequence of the sequenced molecular barcode sequence portion is further specified by the difference. (Eg, each fixed base can be designed to be any one base selected from the group consisting of A, T, G and C; or a combination of A and T, a combination of A and G) Combination, A and C combination, T and G combination, T and C combination, G and C combination, A and T and G combination, A and T and C combination, A and G and C combination From the group consisting of the combination of, and the combination of T, G, and C It may be designed to be a base selected at random from bases contained in any one of the combinations to be-option}.

FIG. 1 is a diagram for explaining a digital quantification method for nucleic acid molecules and its effectiveness. (A) Panel A of FIG. 1 shows a digital counting scheme. A unique molecular barcode is added to each target nucleic acid molecule (add unique molecular barcode). After amplification, both the nucleic acid portion of interest and the barcode portion are sequenced. The number of copies is determined by the number of unique barcodes, not the number of reads. The dotted frame shows the experimental design used in this example. (B) Panel B of FIG. 1 illustrates the first requirement for accurate digital counts: each nucleic acid molecule of interest must be labeled with a different barcode. If the number of unique barcodes measured becomes constant as the diversity of barcode sequences increases, then the diversity of barcode sequences in that range meets the first requirement. (C) Panel C of FIG. 1 illustrates the second requirement for accurate digital counting: all barcode sequences (at least one read) added to the nucleic acid molecule of interest must be detected. It is a figure. If the number of unique barcodes measured becomes constant as the sequence depth is increased, then the sequence depth in that range meets the second requirement. FIG. 2 shows the observed intrinsic features of digital counts with random base barcodes applied to two requirements for accurate digital quantification. (A) Panel A of FIG. 2 shows the dependence of the number of detected clusters (unique bar code shown in gray) on the number of random bases (base length). The results of ST1 are shown. Gray lines indicate the number of unique barcodes. The blue line indicates the number of observed clusters after clustering (Distance = 3). The green line indicates the number of clusters after clustering (Distance = 3) and fixed base match filtering (fixed base number = 6). (B) Panel B of FIG. 2 shows the dependence of the number of barcode clusters on the number of random bases. The yellow line indicates the number of clusters for ST1 having index A and index B after clustering (Distance = 3) and fixed base match filtering (fixed base number = 6). The red dotted line shows the number of barcode clusters for all short templates with index A and index B after clustering (Distance = 3) and fixed base match filtering (fixed base number = 6). (C) Panel C of FIG. 2 shows the dependence of the number of detected molecules on the sequence coverage. Gray lines represent the dependence of the number of unique barcode sequences observed on the number of random bases. The blue line indicates the number of observed clusters after clustering (Distance = 3). The green line indicates the number of clusters after clustering (Distance = 3) and fixed base match filtering (fixed base number = 6). The yellow line indicates the number of barcode clusters for ST1 with index A after clustering (Distance = 3), fixed base match filtering (fixed base number = 6), and exclusion of miss index (eg contaminated index). Show. The red dotted line indicates the bar code cluster of ST1 with index A after clustering (Distance = 3), fixed base match filtering (fixed base number = 6), and misindex (eg, contamination index) and misidentification exclusion. Indicates a number. Error bars indicate standard deviation (n = 8). FIG. 3 shows the results of analysis using Distance and fixed base. In FIG. 3, the results of ST1 with index A (indicated by a circle) and index B (indicated by a triangle) are shown. The random base barcode length was 24. (A) Panel A of FIG. 3 shows the effect of clustering at different Distances on the number of clusters. (B) Panel B of FIG. 3 shows the dependence of the positions of fixed bases at Distance = 3. Asterisk indicates no filtering. Filtering was done using only one fixed base. (C) Panel C of FIG. 3 shows the dependency of the number of fixed bases at Distance = 3. Asterisk indicates no filtering. Filtering was done using the base furthest from the sequence primer site. Figure 4 shows the absolute counts for each template. (A) Panel A of FIG. 4 shows the number of clusters determined for index A as a function of coverage. Distance = 2, fixed base number = 4, random base number = 20. Also, the effects of misindexes (eg, mixed indexes) and misidentifications were excluded. The starting copy number of each template sequence is shown in parentheses. (B) Panel B of FIG. 4 shows the dependence of the number of detected molecules on the number of random bases. Distance = 2, fixed number of bases = 4. Randomly sampled 10% of all reads (coverage for short templates was 13.4-20.3 and coverage for long templates was 12.6-20.9). (C) Panel C of FIG. 4 shows the correlation between the input (ie, the number of molecules before PCR amplification, x-axis) and the output (ie, the result of digital counting, y-axis). The number of outputs was determined from Figures 4A and 11 with a coverage of 12.6 to 20.9 indicated by large symbols. The gray line shows a regression line with a slope of 1 on a logarithmic scale. Circles and triangles correspond to index A and index B, respectively. The Pearson product moment correlation coefficient r and the coefficient of determination R ^{2 of the} linear regression are shown. Error bars indicate standard deviation (n = 8). FIG. 5 shows the required number of random bases for digital counting. (A) In panel A of FIG. 5, the x-axis shows the number of inputs of the molecule to be measured, and the y-axis shows relative values of 0.95 for each curve in panel B of FIG. 4 and panel B of FIG. The number of random bases when reaching the number of target clusters is shown. (B) Panel B of FIG. 5 shows the dependence of the number of barcode clusters on the number of random bases. Clustering (Distance = 2), all templates having indexes A and B after fixed base match filtering (fixed base number = 4), 10% of all reads were randomly sampled (for example, the coverage rate is 12.6 to 20.9). The color corresponds to the sample represented by the plot in panel A of FIG. Error bars represent standard deviation (n = 8). FIG. 6A shows a conventional method involving amplification, indexing, mixing and sequencing. In FIG. 6A, the barcode sequence is not used and a unique index is added to the amplified sequence for each sample, mixed and sequenced. The index may be added before amplification. FIG. 6B shows a conventional method involving amplification, indexing, mixing and sequencing. Here, in the conventional method, misindex addition may occur, but the misindex that occurred cannot be identified. FIG. 6C illustrates the use of molecular barcodes. A unique barcode sequence is labeled for each nucleic acid sequence of interest having sequence 1, and each molecule is uniquely labeled. FIG. 6D shows the use of molecular barcodes including molecular barcode addition, amplification, indexing, mixing and sequencing. A scheme for adding a unique barcode to a nucleic acid molecule and a unique index for each sample, and mixing and sequencing a plurality of samples is shown. The index may be added after the addition of the molecular barcode and before the amplification. FIG. 6E shows the use of molecular barcodes including molecular barcode addition, amplification, indexing, mixing and sequencing. The scheme of an example of the identification method of the misindex in the 1st Embodiment of this invention is shown. Misindexes may occur, but in the first embodiment of the invention, the misindexes that occur can be identified. Amplification products from the same molecule have the same index, and it is assumed that the number of misindexed molecules is less than the number of correctly indexed molecules. In the first embodiment of the present invention, when two types of indexes are added to the same barcode array, the most frequent index in the number of reads is determined as the correct index. FIG. 7A shows a scheme for adding a barcode to a nucleic acid molecule of interest contained in multiple samples. FIG. 7B shows an indexing and amplification scheme, showing the case where partial switching occurs due to indexes contaminated with other indexes. FIG. 7C shows counting of the number of barcodes, confirmation of the same barcode, and identification of an error (mispair between index and barcode). A scheme for identifying a molecule with a small read number (copy number) as a misindex when different indexes A and B are added to the same barcode sequence is shown. FIG. 8 is a supplementary drawing of FIG. 2 and shows the observed intrinsic features of digital counting using barcodes with random bases. (A) Panel A of FIG. 8 shows the dependence of the number of detected clusters on the number of random bases (base length) for ST1, ST2, LT1 and LT2. Gray lines represent the dependence of the number of unique barcode sequences observed on the number of random bases. The blue line indicates the number of observed clusters after clustering (Distance = 3). The green line indicates the number of clusters after clustering (Distance = 3) and fixed base match filtering (fixed base number = 6). (B) Panel B of FIG. 8 shows the dependence of the number of barcode clusters on the number of random bases. Yellow lines indicate clustering (Distance = 3), fixed base match filtering (fixed base number = 6 for ST2, fixed base number = 12 for LT2), and misindex (for example, misindex of contamination). The number of clusters for ST2 (top panel) and LT2 (bottom panel) with indices A and B after exclusion is shown. The dark yellow line is LT1 with index A and index B (lower panel) after clustering (Distance = 3), fixed base match filtering (fixed base number = 12), and exclusion of miss index (for example, contamination index). Shows the number of clusters for. Red dotted lines are all long lines with index A and index B after clustering (Distance = 3), fixed base match filtering (fixed base number = 6), miss index (eg contaminated index) and misidentification exclusion. The number of barcode clusters for the template is shown. (C) Panel C of FIG. 8 shows the dependence of the number of detected molecules on the sequence coverage for ST1, ST2, LT1 and LT2. In FIG. 8 panel C, the color of the lines is the same as in panels A and B of FIG. 8 above, except for dark yellow. Error bars indicate standard deviation (n = 8). FIG. 9 is a supplementary drawing of FIG. 3 and shows the analysis results of ST2, LT1 and LT2 using Distance and a fixed base. (A) Panel A of FIG. 9 is the same as panel A of FIG. 3, but shows the effect of clustering with various Distance parameters on the number of clusters for ST2, LT1 and LT2. (B) Panel B of FIG. 9 is the same as panel B of FIG. 3, but shows the dependence of fixed base position on ST2, LT1 and LT2. Asterisk indicates no filtering. (C) Panel C of FIG. 9 is the same as panel C of FIG. 3, but shows the dependence of the number of fixed bases on ST2, LT1 and LT2. FIG. 10 shows a histogram of the number of reads in each cluster for ST1 with index A (see panel A) and index B (see panel B). The colors correspond to the sample colors in the plot of Figure 2C. FIG. 11 is a supplementary drawing of FIG. 4, showing absolute counts for each mold. Same as panel A in FIG. 4, but for index B (indicated by triangles), the determined number of clusters versus coverage is shown. Error bars indicate standard deviation (n = 8). FIG. 12 is a supplementary drawing of FIG. 5, showing an estimation of the required number of random bases for digital counting. This plot is the same as FIG. 5A except that it is a linear regression (R ² = 0.971) on a logarithmic scale up to 38 random bases (the longest in this example). The equation for the regression curve above in this case was y = 2.2 * log (x) +5.5. FIG. 13 shows the barcode design and the number of molecular inputs. Uppercase letters in the sequences of LT1 to 6 are binding sites of PCR amplification primers. The barcode indicates a random region containing random bases and fixed bases, and the target indicates a target nucleic acid. LT1-6 were PAGE-purified products, and the 5'ends of ST1-5 were amine-modified. Fixed bases between random bases help avoid long homopolymer barcodes that may have lower amplification efficiencies. N indicates either A, T, G, or C. Figure 14 shows the primer sequences for library preparation. The underlined portion indicates the index array (the index A is included in Rv primer1, and the index B is included in Rv primer2). All primers were PAGE purified products. FIG. 15 shows the number of leads in each process. * This fraction can be larger than the number of reads in the exclusion of missed indexes (eg mixed indexes) (see examples).

Detailed explanation of the invention

As used herein, a “molecular barcode” is a tag having a unique sequence added to a nucleic acid molecule on a molecule-by-molecule basis. It is also called “primer ID” and “unique molecular identifier (UMI)”. When a molecular barcode having a unique sequence different for each molecule is added to the nucleic acid molecule, the number of molecules of the nucleic acid contained in the sample before being subjected to treatment such as amplification is added. It will be possible to make a digital (or qualitative) decision based on the number of barcode types. This method for determining nucleic acid molecules has come to the forefront with the development of a next-generation sequencer platform that enables the analysis of a large amount of nucleic acid sequences in a single run. Various methods have been developed to digitally determine the number. This method for determining the number of nucleic acid molecules is capable of digitally counting the number of molecules as the number of barcode types (sometimes referred to as "the number of unique barcodes"). , Etc. This digital counting method can digitally accurately determine the absolute number of molecules in a sample even in the presence of noise and bias in the measurement system. The most widely used application of this digital counting method is RNA-Seq using a molecular barcode, that is, digital RNA-Seq (dRNA-Seq) or quantitative RNA-Seq. Since dRNA-Seq functions well even in a small amount of sample, it is often used for single-cell gene expression analysis.
Digital counting methods are also used in many applications in next-generation sequencer platforms that can acquire large amounts of sequence data. Examples of such applications include, in addition to RNA-Seq, single-nucleotide resolution UV cross-linking immunoprecipitation (iCLIP), antibody repertoire analysis, gene analysis of bacterial 16S rRNA, And exonucleases, unique barcodes and nucleotide resolution through exonuclease, unique barcode and single ligation (ChIP-nexus: chromatin immunoprecipitation experiments with nucleotide resolution through exonuclease, unique barcode and single ligation).
In this digital counting method, by using a sufficient number of types of molecular barcodes for the total number of nucleic acid molecules present in the sample, the same barcode is applied to multiple nucleic acid molecules present in the original sample. The possibility of being added is substantially limited, which allows the number of types of barcode sequences to be related to the number of nucleic acid molecules present in the sample. In this way, the quantification of nucleic acid molecules present in a sample is possible by using a molecular barcode containing nucleotide sequences of sufficient diversity. The molecular barcode can be obtained, for example, as a nucleic acid group having random bases. Since the molecular barcode is focused on the number of kinds of sequences to determine the number of molecules to be measured, the sequences are random (the sequences are diverse, and humans do not need to recognize the contents of the sequences). It can be said that it may be the one that has been synthesized to. Alternatively, the molecular barcode may be a group of nucleic acids of known sequence designed to provide sufficient diversity. In the present specification, the molecular barcode may be simply referred to as a barcode, and the sequence possessed by the molecular barcode may be referred to as a barcode sequence. As used herein, the number of unique barcode sequences is a number that represents the degree of diversity of the barcode sequences. The number of unique barcode sequences will be n if n different barcode sequences are detected, where n is a natural number. As used herein, the number of random bases means the base length of random bases. As used herein, a random base means a continuous base having a random sequence. Random bases may consist of two bases, three bases, or four bases.

  As used herein, an "index" is a nucleic acid that is a unique label attached to a nucleic acid molecule for each sample from which it is derived. For example, an index having a different nucleotide sequence may be added to each sample. By adding the same index to all nucleic acid molecules derived from a certain sample, when multiple samples are mixed and sequenced, the sample from which each nucleic acid molecule is derived is determined based on the sequence of the added index. Can be specified. Because of the large capacity of a single sequence on the next-generation sequencer platform, multiple samples can be mixed and sequenced in a single run, and the index is useful, for example, in such cases. is there. The index may be added before, during, or after treatment (eg, amplification) of the nucleic acid molecule.

  In the present specification, “template”, “target nucleic acid”, “target nucleic acid molecule”, “target nucleic acid” or “target nucleic acid molecule” means a nucleic acid molecule (eg, DNA or RNA) to be quantified in a digital quantification method. ), And can be used interchangeably. In the present specification, a sequence which the nucleic acid molecule of interest originally has (that is, a sequence before adding a barcode or index for analysis) is referred to as a nucleic acid sequence of interest.

  As used herein, "nucleic acid" means a macromolecule having a nucleic acid sequence. Nucleic acids include deoxyribonucleic acid (DNA) and ribonucleic acid (RNA). Ribonucleic acids include messenger RNA (mRNA), non-coding RNA such as microRNA, transfer RNA (rRNA), and ribosomal RNA (rRNA).

  As used herein, "sequence depth" refers to the total amount or total number of molecules sequenced. For example, higher sequence depths (ie, more sequence information available) may increase the likelihood that low-abundance sequences will be detected in the sample. In the present specification, the “coverage ratio” means the average number of reads of each cluster obtained by clustering as derived from the same nucleic acid molecule (read number / cluster).

  As used herein, the term “unique to each molecule” means that at least some of the molecules included in the system are different from each other. The term “unique to each molecule” means all molecules, substantially all molecules, or most of the molecules contained in the system (eg, 50% or more, 60% or more, 70% or more, 80% or more, 90% or more). % Or more, 95% or more, 96% or more, 97% or more, 98% or more, or 99% or more).

The conventional procedure for digital quantification of nucleic acids is described below (see FIG. 1, panel A).
DNA (molecular barcode) containing various foreign sequences is uniquely added to each of nucleic acid (target nucleic acid molecule) such as RNA molecule or DNA (eg, complementary DNA or cDNA) molecule (ie, nucleic acid). Add a molecular barcode with a different sequence for each molecule) (see, eg, FIG. 6C). Such a nucleic acid to which a molecular barcode having a unique sequence for each molecule is added may be referred to as “barcode added nucleic acid”. The barcode-added target nucleic acid molecule (cDNA obtained from RNA when the starting material nucleic acid is RNA) is amplified (see, for example, FIG. 6D). The nucleic acid sequence of interest and the barcode sequence of the barcode-added and amplified nucleic acid are sequenced in tandem (see, for example, FIG. 6D). As theoretically suggested, for each nucleic acid of interest, the number of unique barcodes added to the nucleic acid sequence of interest, rather than the number of amplified molecules (the so-called "read number"), is quantified and the original The absolute copy number of the nucleic acid molecule of interest (before amplification) can be determined. In this digital quantification method, since the number of kinds of barcode sequences is focused, the barcode sequence may be added to the target nucleic acid molecule so as to have a unique sequence for each nucleic acid molecule. It doesn't matter what it is. In the digital quantification method, a barcode whose specific sequence is known may be used.
The next-generation sequencer platform has developed, and it has become possible to decode a large amount of nucleotide sequences with a single sequencing (run). This has increased the need for sequencing multiple samples in a single run without exhausting the sequencing capabilities in single sample measurements. A unique index can be added to each sample to distinguish which sample the nucleic acid came from, while sequencing multiple samples in one run. According to the present invention, the index may be added to the target nucleic acid molecule so as to have any sequence as long as it is unique for each sample, and it does not matter what the specific sequence is. In the digital quantification method, an index whose specific sequence is known may be used.
According to the present invention, the index may be added to the amplified target nucleic acid molecule after the target nucleic acid molecule is amplified, or may be added to the target nucleic acid molecule before the target nucleic acid molecule is amplified. May be done. The index may be added after amplification is performed on each sample. For example, the index can be added to each amplification product by adapter ligation. Alternatively, the index may be added while the nucleic acid molecule of interest is amplified. For example, indexing can be done during amplification of the nucleic acid molecule by inclusion in the sequence of the primer.
In the present invention, when the index is added to the nucleic acid molecule of interest before amplification, the index may be added to the nucleic acid molecule of interest before, simultaneously with, or after the addition of the barcode sequence. The index, barcode sequence, and nucleic acid molecule of interest may be linked in any order. The index may be provided in association with the barcode array. When a target nucleic acid molecule contained in a specific sample is subjected to digital quantification using a molecular barcode, the target nucleic acid molecule derived from the specific sample can be specified by using the index as an index and added to the target nucleic acid sequence. The number of different types of barcode sequences (the number of unique barcodes) generated is quantified and the absolute copy number of the original (prior to amplification) nucleic acid molecule of interest is determined (see, eg, Figure 6D).

According to the present invention, in a method for digitally quantifying a target nucleic acid molecule using an index and a barcode, when the target nucleic acid molecule is quantified by mixing a plurality of samples, the index is a nucleic acid derived from a different unexpected sample. It has become clear that a problem of being added to the above may occur (see FIG. 6E and FIG. 7B). This problem can occur when an index is used, and is called index switching, index hopping, misindexing, or the like. The existence of index switching problems has already been pointed out (Sinha, R. et al. Index switching causes “spreading-of-signal” among multiplexed samples in Illumina HiSeq 4000 DNA sequencing. Biorxiv, 10.1101 / 125724 (2017)) To date, no effective solution has been reported.
Also according to the present invention, when counting the number of types of barcode sequences, the sequences that should be judged to be the same are different due to mutations (eg, insertion, substitution, and deletion) that occur in the barcode sequences. It has become clear that there may be problems identified as. These problems can occur whether or not indexes are used.

The present invention provides a solution to each of these problems.
In the digital quantification method that uses a unique index for a sample to distinguish between samples, multiple types of indexes are not added to the same barcode for the target nucleic acid molecule to which the barcode and index are added. Can be assumed (because a unique barcode is added to each nucleic acid molecule). On the other hand, in the present invention, when multiple indices are found in the cluster of nucleic acid molecules to which the same barcode is added, it can be determined that a misindex has occurred (for example, FIG. 6E). And FIG. 7C). When multiple indices are found in a cluster of nucleic acid molecules with the same barcode, the number of existing index sequences is compared and the most abundant index sequence is determined by the correctly indexed sequences. Yes (see, eg, FIGS. 6E and 7C). This can accommodate misindexes (eg, by excluding sequences other than the most abundant index sequence in a cluster). It goes without saying that this method can be carried out independently of the sequence of the nucleic acid molecule of interest. Thus, this method may or may not include decoding the nucleic acid sequence of interest. This method corresponds to the first embodiment described below.

  In the digital quantification method, the recognition of the difference between the index sequence and the barcode sequence affects the quantification accuracy. For example, if a barcode sequence is recognized as a different sequence due to a mutation (eg, insertion, substitution, and deletion) of bases in the sequence, whether or not an index is added, the number of types of sequences is determined. Digital quantification, which is used to determine the original number of molecules before being subjected to amplification etc., makes the determination of the number of molecules inaccurate. On the other hand, in the present invention, for substitution of bases in a barcode, sequences included in a certain distance are clustered as one cluster, and the number of molecules is determined based on the number of clusters. It is possible to deal with the problem that the sequences are originally recognized as different sequences due to the substitution of bases. Here, “distance” means the number of bases that differ between two predetermined barcode sequences. For example, if one barcode sequence is exactly the same as another barcode sequence except for one base change at any one position, the distance between these two barcode sequences is 1. is there. Also, for example, the distance between these two barcode sequences is 2 if they are exactly the same except for two base changes at any two positions. For example, if one barcode sequence is exactly the same as another barcode sequence except for three base changes at any three positions, the distance between these two barcode sequences is 3 Is. It is believed that the greater the diversity of barcode sequences, the greater the accuracy of the method of the first embodiment. The value of the distance is not particularly limited as long as it is appropriately determined according to the present disclosure, but is, for example, 1 to 10, preferably 1 to 5, more preferably 1 to 3, and further preferably 3. It goes without saying that this method can be carried out independently of the sequence of the nucleic acid molecule of interest. Thus, this method may or may not include decoding the nucleic acid sequence of interest. This method corresponds to the second embodiment described below. In the system to which the index is added, it can be used similarly when determining the difference between the indexes.

  In addition, for example, for insertions or deletions of bases in a barcode sequence (insertions and deletions may be collectively referred to as "indel"), regardless of whether an index is added or not, the barcode By making the base at the fixed position of the fixed base as the fixed base (that is, the base at the predetermined position in the barcode sequence is specified or the specified base), the indel is generated with the fixed base not existing at the predetermined position Can be detected (this method is sometimes referred to herein as "fixed base match filtering"). That is, it is determined that insertion or deletion of a base has occurred in the barcode sequence when the sequence includes a base different from the original base at any of the fixed base positions in the sequence. The number of fixed bases in the barcode sequence may be appropriately determined according to the present disclosure and is not limited, but is, for example, 1 to 15, preferably 2 to 12, more preferably 3 to 10, and further preferably 4 to. There are six. It goes without saying that this method can be carried out independently of the sequence of the nucleic acid molecule of interest. Thus, this method may or may not include decoding the nucleic acid sequence of interest. This method corresponds to the third embodiment described below. In the system to which the index is added, it can be used similarly when determining the difference between the indexes.

  Hereinafter, each of the first embodiment, the second embodiment, and the third embodiment will be described. It should be noted that these embodiments can be implemented in combination, and the present invention encompasses such a combination of possible embodiments. The embodiments below include non-limiting examples of combined embodiments.

A first embodiment of the present invention That is, according to the first embodiment of the present invention,
Sequencing using a mixture of a plurality of samples containing a target nucleic acid molecule having a unique index for each sample containing a plurality of nucleic acid molecules and a unique or arbitrary molecular barcode for each nucleic acid molecule (multiplex sequencing) From the sequence information obtained, a method for determining the correct pair or mispair of the index and the molecular barcode added to the target nucleic acid molecule,
(E) From the obtained sequence information, a sequence having a specific index or a sequence similar thereto, a sequence having a specific molecular barcode or a sequence similar thereto, or a sequence containing a target nucleic acid molecule or a sequence similar thereto Selecting sequences and creating groups with the selected sequences;
(F) In the group created in (E) above, the pair of the index and the molecule barcode with the highest detection frequency is determined as the correct pair of the index and the molecule barcode, and / or the index and the molecule with the low detection frequency are determined. A pair of barcodes (for example, a pair having a detection frequency lower than a certain reference value, and the certain reference value is 99.5% or less, 99% or less, 90% or less, 80% or less, 70% or less, 60% or less in a group. Examples include, but are not limited to, values of% or less, 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less. A pair of frequencies), and determining at least any one or all of them as mispairs of the index and the molecular barcode.
A method is provided, including:

In a first embodiment of the invention, the method of the invention comprises
(A) separately obtaining a plurality of samples containing nucleic acid molecules (for example, DNA or RNA), and {at least one of the samples contains a target nucleic acid molecule},
(B) {for example, in each of the obtained plurality of samples}, before amplifying the nucleic acid molecule contained in the sample, each nucleic acid molecule of interest is linked with a unique or arbitrary molecular barcode of each nucleic acid molecule, Obtaining a target nucleic acid molecule in which different molecular barcodes are linked,
(C) {For example, before mixing a plurality of samples} A target nucleic acid molecule in which a unique index is added to a target nucleic acid molecule for each sample containing a plurality of target nucleic acid molecules and a different index is linked to each derived sample And the step (C) may be carried out after the step (B), or the step (B) may be carried out after the step (C); and the step (B) or (C). A), the nucleic acid molecule can be amplified to obtain an amplification product of the target nucleic acid molecule),
(D) In a mixture containing the amplification products of nucleic acid molecules obtained after (B) and (C) above (the sample is mixed after the step (C), and the sample is mixed after the step (B). ) May be carried out, and all the samples may be mixed after carrying out step (B), and the amplification product of the nucleic acid molecule linked with the molecular barcode is obtained after step (B), The sample may be mixed before obtaining the amplification product, and the sample containing the amplification product may be mixed after obtaining the amplification product), the index unique to each sample and the unique or arbitrary to each nucleic acid molecule of interest. Sequence of the nucleic acid molecule to which the molecular barcode is added to determine the sequence of the index portion and the sequence of the molecular barcode portion and the sequence of the target nucleic acid molecule portion linked to the nucleic acid molecule sequence for each nucleic acid molecule. When,
May be further included.

  In the first embodiment of the present invention, as the index, any index can be used as long as it has a unique base sequence for each sample. The index may have a given sequence (for example, it may be possible to determine which sample it came from by referencing the sequence), but the sequence is unknown. (For example, it is not possible to determine from which sample the sample is derived by referring to the sequence, and it may be possible to know that the sample is derived from a different sample due to the different sequence).

In the first embodiment of the present invention, the molecular barcode can be made to have sufficient diversity with respect to the number of nucleic acid molecules in the sample. The molecular barcode may have any base sequence as long as it has sufficient diversity with respect to the number of nucleic acid molecules in the sample. The sequence of the molecular barcode can be a randomly determined sequence (randomly determined sequence) for the purpose of saving the effort of designing the sequence. For example, the molecular barcode may have the above sufficient diversity by including a plurality of randomly determined bases (that is, random bases). In order to secure the diversity of the molecular barcode, the length of the nucleotide sequence of the molecular barcode can be increased. When random bases are used in the digital quantification of target nucleic acids having a predetermined diversity, the number of random bases in the base sequence of the required molecular barcode is experimentally determined based on the graph illustrated in FIG. You may. Although not limiting the present invention, for example, by setting the number of random bases in the base sequence of the molecular barcode to be 38 or more, sufficient diversity is secured to digitally quantify the number of molecules up to 10 ^15. It can be understood from the examples that what can be done. If four bases are randomly arranged and the base length is 38, the diversity of the molecular barcode theoretically reaches 4 ³⁸ (that is, about 7.56 × 10 ²² ). The number of random bases in the molecular barcode is, for example, 6 or more, 7 or more, 8 or more, 9 or more, 10 or more, 11 or more, 12 or more, 13 or more, 14 or more, 15 or more in order to secure sequence diversity. , 16 or more, 17 or more, 18 or more, 19 or more, or 20 or more. Alternatively, the number of random bases may be 25 or more, 30 or more, 35 or more, 40 or more.

  In the first embodiment of the present invention, the plurality of samples are 2 or more, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more, 9 or more, or 10 or more samples, and are It is a distinguishable number, but there is no upper limit to the number.

  In the first embodiment of the present invention, in the above (E), a sequence having a specific index, a sequence having a specific molecular barcode, and / or a sequence containing a nucleic acid molecule of interest is based on the sequence identity. Groups can be formed by the selected sequences. Here, by selecting a sequence having a specific molecular barcode and forming a group for each molecular barcode by the selected sequence, it is possible to form a number of groups corresponding to the number of types of molecular barcodes. . In addition, by selecting an array having a specific index and forming a group for each index by the selected array, the number corresponding to the number of indexes (for example, the number of samples when a different index is added to each sample) Can form a group of. Further, by selecting a sequence having a specific target nucleic acid and forming a group with the selected sequence, a nucleic acid group containing the target nucleic acid can be obtained.

In the first embodiment of the present invention, in the above (E), the step of forming groups has the same sequence {preferably in the sequence of the molecular barcode portion} based on the sequence identity or similarity. (For example, the sequence may be changed by any of the steps (A) to (D)), the clustering is performed as a group to create a group.
In the first embodiment of the present invention, for example, the above (E) can be carried out in combination with the second embodiment. Details will be described in the second embodiment.
Furthermore, in the first embodiment of the present invention, for example, the above (E) can be implemented in combination with the second embodiment and the third embodiment. Details will be described in the third embodiment.

  In the first embodiment of the present invention, in the above (F), for each of the groups created in the above (E), the pair of the index and the molecular barcode with the highest detection frequency is set as the correct pair of the index and the molecular barcode. You can decide. In the first embodiment of the present invention, at least any one or all of the pairs of the index and the molecular barcode having a low detection frequency can be determined as the mispair of the index and the molecular barcode. In the first embodiment of the present invention, the pair of the index and the molecular barcode having the highest detection frequency is determined as the correct pair of the index and the molecular barcode, and at least the pair of the index and the molecular barcode having a low detection frequency is determined. Any one or all can be determined as mispairs of the index and molecular barcode. In the first embodiment of the invention, nucleic acid molecules determined to be mispairs can be excluded from counting the number of molecules. The determination of the correct pair and the determination of the mispair can be performed independently of the sequence of the nucleic acid molecule of interest. For example, the target nucleic acid molecule may be selected and then the correct pair and Mispair may be determined respectively; or, the correct pair and Mispair may be determined and then the target nucleic acid molecule may be selected. You may.

For example, in one aspect, in the above (E), when a sequence having a specific molecular barcode is selected to create a group for each molecular barcode,
(i) In step (F), it is possible to determine the pair of the index and the molecular barcode having the highest detection frequency among the created groups as the correct pair of the index and the molecular barcode; or
(ii) A pair of an index and a molecular barcode having a low detection frequency in the group prepared in step (F) (for example, a pair having a detection frequency lower than a certain reference value, and the certain reference value is a group). Examples include, but are not limited to, values of 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less. May be present) can be determined to be at least one and / or all mispairs of the index and the molecular barcode.

For example, in one aspect, in (E), when an array having a specific index is selected and a group is created for each index,
(iii) In the step (F), regarding the nucleic acid sequences having the molecular barcodes appearing in a plurality of groups, the barcode-index pair having the largest number of reads is determined as the correct barcode-index pair, or The most frequently detected pair of index and molecular barcode can be determined as the correct pair of index and molecular barcode; or
(iv) In the step (F), for a sequence having a molecular barcode that appears in a plurality of groups, a pair of an index and a molecular barcode having a low detection frequency (for example, a pair having a detection frequency lower than a certain reference value, The constant reference value includes, but is not limited to, values of 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% or less in the group. The second and subsequent detection frequency pairs may be determined to be at least one or all of the indexes and the mispair of the molecular barcode.

For example, in one embodiment, in the above (E), when a sequence containing the target nucleic acid molecule is selected to create a group,
(v) In step (F), a sequence having a specific index is further selected from the group to form a subgroup, and the nucleic acid sequence having a molecular barcode that appears in a plurality of subgroups has the highest number of reads. The barcode and index pair can be determined as the correct barcode and index pair, or the most frequently detected index and molecular barcode pair can be determined as the correct index and molecular barcode pair;
(vi) In step (F), a molecule having a specific molecular barcode is further selected from the group to create a subgroup, and the index and the molecular barcode with the highest detection frequency in one created subgroup are selected. The pair can be determined as the correct pair of index and molecular barcode;
(vii) In the step (F), a molecule having a specific index is further selected from the group to form a subgroup, and a nucleic acid molecule having a molecular barcode that appears in a plurality of subgroups has an index with a low detection frequency. A pair of molecular barcodes (e.g., a pair having a detection frequency lower than a certain reference value, and a certain reference value is 50% or less, 40% or less, 30% or less, 20% or less, 10% or less in a group, 5% or less, 1% or less, but not limited to these. For example, it may be a pair of detection frequencies from the second onward). Can be determined to be a mispair; or
(viii) In step (F), a molecule having a specific molecular barcode is further selected from the group to create a subgroup, and a pair of an index and a molecular barcode having a low detection frequency in one created subgroup ( For example, a pair with a detection frequency lower than a certain reference value, and the certain reference value is 50% or less, 40% or less, 30% or less, 20% or less, 10% or less, 5% or less, 1% in the group. The following values are included, but not limited to these, and may be, for example, the second and subsequent detection frequency pairs.) At least any one or all of them are determined as mispairs of the index and the molecular barcode. You can

  In this way, in the first embodiment of the present invention, the correct pair of barcode array and index array can be determined and / or the mispair can be determined. As shown in Examples described later, by not counting mispairs, the accuracy of digital quantification of the target nucleic acid molecule can be improved.

Second Embodiment of the Present Invention In a digital quantification method of a nucleic acid molecule using a barcode sequence, a mutation (insertion, substitution or deletion) occurs in the barcode sequence during analysis, and the mutation has a quantification accuracy. It has been revealed that The second embodiment of the present invention is a method for digitally quantifying a target nucleic acid molecule using a barcode sequence, and based on the information of the obtained nucleic acid sequence, the sequence possessed by the molecular barcode after mutation has sequence similarity. And clustering together with the sequences of (clustering). This seeks to minimize the effects of mutations within the barcode sequence that occur during analysis. In the second embodiment, for example, in an environment in which a sequence similar to a molecular barcode is unlikely to be contained, the similar sequence may be generated by mutation (insertion, substitution, or deletion) from the same sequence. It is suggested that this clustering improves the accuracy of digital quantification even in the practical examples.
More specifically, for example, the step of forming a group had an identical sequence {preferably in the sequence of the molecular barcode portion} having the same sequence determined on the basis of sequence identity or similarity {eg, the step When (A) to (D) are carried out, the sequence may be mutated by any of these steps.} It may be possible to create a group by clustering the putative molecules. Therefore, a sequence having similarity with a sequence having a specific index includes a sequence having a specific index and a sequence having similarity with a sequence having a specific index.

In the second embodiment of the present invention, for example, in a digital quantification method of a nucleic acid molecule of interest using a barcode sequence, the obtained nucleic acid sequence is subjected to an index, barcode and / or nucleic acid molecule of interest based on sequence similarity. Can be divided into groups (clustering). In an aspect of the second embodiment of the present invention, for example, the clustering is
(I) In the sequence of the molecular barcode portion, the nucleic acid molecule group having the same sequence as the unique molecular barcode sequence is classified into the same cluster (that is, Distance = 0);
(Ii) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a unique molecular barcode sequence and a sequence having a mismatch of up to 1 base is classified into the same cluster (that is, Distance = 1);
(Iii) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a unique molecular barcode sequence and a sequence having a mismatch of up to 2 bases is classified into the same cluster (that is, Distance = 2); Or (iv) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a sequence having a mismatch of up to 3 bases with the sequence of the unique molecular barcode is classified into the same cluster (that is, Distance = 3) . By doing so, the artificial increase in the kinds of nucleic acid sequences due to the mutation of 0 to 3 bases that can occur in the digital quantification method is corrected.
This aspect of the second embodiment, when combined with the first embodiment, can be implemented in step (E) above.

In the second embodiment of the present invention, for example, in a digital quantification method of a nucleic acid molecule of interest using a barcode sequence, the obtained nucleic acid sequence is subjected to an index, barcode and / or nucleic acid molecule of interest based on sequence similarity. Can be divided into groups (clustering). In an aspect of the second embodiment of the present invention, the clustering is performed by, for example, inserting or deleting bases (eg up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of the molecular barcode portion. ) Are grouped into the same cluster.
This aspect of the second embodiment, when combined with the first embodiment, can be implemented in step (E) above.

In the second embodiment of the present invention, for example, in a digital quantification method of a nucleic acid molecule of interest using a barcode sequence, the obtained nucleic acid sequence is subjected to an index, barcode and / or nucleic acid molecule of interest based on sequence similarity. Can be divided into groups (clustering). In an aspect of the second embodiment of the invention, the clustering is performed by, for example, inserting or deleting bases (eg up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of the molecular barcode portion. ) Is performed on the nucleic acid molecule group obtained by excluding the sequence sequence.
This aspect of the second embodiment, when combined with the first embodiment, can be implemented in step (E) above.

  Further, for example, in an aspect of the second embodiment, a nucleic acid sequence can be selected depending on whether or not a particular barcode sequence is similar, and a group can be created by the selected sequence. Here, “similar” means that the sequences differ by 1 base, 2 bases, 3 bases or more (for example, insertion, deletion or substitution), but the other bases are the same. The ratio of matching bases between similar base sequences is, for example, 50% or more, 55% or more, 60% or more, 65% or more, 70% or more, 75% or more, 80% or more, 85% or more, 90% or more. , 95% or more, 96% or more, 97% or more, 98% or more, or 99% or more.

Third Embodiment of the Invention In the digital quantification method of a target nucleic acid molecule using a barcode sequence, insertion or deletion (indel) may occur in the obtained nucleic acid sequence. In a third embodiment of the invention, in the detection of indels that may occur for nucleic acid sequences (particularly barcode sequences), one or more (eg one Whether some (one or more) or all of the two, three, four, five, or six or more fixed bases are changed to a base other than the predetermined fixed base at the original position. Can be detected. In a third embodiment of the invention, also in the detection of indels that may occur for nucleic acid sequences (particularly barcode sequences) one or more (eg 1 One, two, three, four, five, or six or more) positions of each fixed base, and the respective positions of one or more fixed bases in the sequence of the sequence-decoded barcode sequence portion. May further include identifying by comparing the target positions {eg, each fixed base is usually designed to be any one base selected from the group consisting of A, T, G and C. Or each fixed base can be a combination of A and T, a combination of A and G, a combination of A and C, a combination of T and G, a combination of T and C, a combination of G and C, A and T. And G, A, T and C Can be designed to be a base selected from the bases included in any one combination selected from the group consisting of a combination of A, G and C, and a combination of T, G and C} . As a result, one or more fixed bases are present at positions deviated from a predetermined position as an index, and preferably other bases are present at positions where fixed bases should be present as an additional index. Can be detected. For example, if one or more, for example, two or more fixed bases are present at positions deviating from the predetermined position by the same number of bases, it can be determined that indel has been detected. When indel is detected, a group of nucleic acid molecules having a sequence sequenced as having indel may be classified into the same cluster as a sequence having no indel, or a group of nucleic acid molecules having a sequence sequenced as having indel May be excluded (for example, the nucleic acid molecule group having a sequence sequenced as having indel may be excluded from the obtained sequence information, or the nucleic acid molecule group having a sequence sequence having as indel may be excluded. Nucleic acid molecule groups may be excluded and clustered). In this aspect, when there are two or more fixed bases, the fixed bases may preferably have one or more other bases interposed between the fixed bases. Here, the term “fixed base” means that a plurality of barcode sequences have a common position at a predetermined position from the ends (5 ′ end, 3 ′ end, or 5 ′ end and 3 ′ end) of the barcode sequence. (In this case, the common base may be a base determined by a common design among a plurality of barcode sequences as described above).
This aspect of the third embodiment, when combined with the first embodiment, can be implemented in step (E) above. This aspect of the third embodiment, when combined with the second embodiment, can be implemented in indel detection.

In addition, the first embodiment of the present invention is
A method for analyzing nucleic acids:
(I) subjecting a mixture of a plurality of target nucleic acid molecules having a molecular barcode and an index to sequencing to obtain sequence information;
(II) A sequence having a specific index or a sequence similar thereto and / or a sequence having a specific molecular barcode or a sequence similar thereto is selected from the sequence information obtained in (I) above, and selected. Creating a group with the arranged array,
(III) in the group created in (II) above, determining the pair of the index and the molecular barcode with the highest detection frequency as the correct pair of the index and the molecular barcode,
May be included.

Further, a second embodiment of the present invention is a method for analyzing nucleic acid, which comprises:
(I) subjecting a mixture of a plurality of nucleic acid molecules having a molecular barcode to sequencing to obtain sequence information;
(II) a step of selecting a sequence having a specific molecular barcode or a sequence similar thereto from the sequence information obtained in (I) above, and creating a group by the selected sequence;
May be included.

Further, a third embodiment of the present invention is a method for analyzing nucleic acid, which comprises:
(I) subjecting a mixture of a plurality of nucleic acid molecules having a molecular barcode having a fixed base at a specific position to sequencing to obtain sequence information;
(IIa) excluding from the analysis a sequence having a molecular barcode that does not include the fixed base at the specific position,
May be included.

  In each of the first, second, and third embodiments, the nucleic acid molecule of interest to which at least the molecular barcode is added may be subjected to amplification before step (I). Here, the target nucleic acid molecule to which at least the molecular barcode is added means that the index may be further added as long as at least the molecular barcode is added, and the index may not be added. To do.

In each of the first, second, and third embodiments described above, the molecular barcode is a well-known method, for example, when a nucleic acid molecule of interest is amplified using a primer containing a molecular barcode sequence (eg, polymerase). It can be added to the nucleic acid molecule of interest (by chain reaction).
In each of the first, second, and third embodiments, the index may be added to the amplification product of the target nucleic acid molecule to which the molecular barcode is added. As a method for adding an index to the amplification product, a well-known method, for example, an adapter ligation method using an adapter having an index sequence can be mentioned.
In each of the first, second, and third embodiments above, the index may be added to the nucleic acid molecule of interest along with the molecular barcode. For example, as a method for adding an index and a molecular barcode to a target nucleic acid molecule, a well-known method, for example, amplification of the target nucleic acid molecule using a primer containing the index and the molecular barcode sequence (eg, polymerase chain reaction) There is a method.

  The method of the first embodiment can be implemented in combination with the second embodiment. For example, in each of the above-mentioned first and second embodiments, a sequence similar to the sequence having the specific molecular barcode in step (II) is equal to or less than the sequence having the specific molecular barcode in a predetermined number of bases. It may be a sequence containing the mismatched bases in the molecular barcode sequence part. Here, the predetermined number of bases is an integer in the range of 1 to 10, 1 to 9, 1 to 8, 1 to 7, 1 to 6, 1 to 5, 1 to 4, 1 to 3, or 1 to 2. , Or 0, 1, 2, or 3. In the sequence containing a mismatch base of a predetermined number of bases or less in the molecular barcode sequence portion, the bases other than the mismatch bases exactly match the sequence of the specific molecular barcode.

The method of the first embodiment can be carried out in combination with the third embodiment. Moreover, the method of the second embodiment can be implemented in combination with the method of the third embodiment.
For example, in each of the first and second embodiments, the molecular barcode may have a fixed base at a specific position.
In each of the first and second embodiments, a sequence similar to the sequence having the specific molecular barcode in step (II) contains the fixed base at the specific position, and / or the fixed base. Position may be selected based on the position being shifted from the particular position.
Each of the first and second embodiments may further include excluding from the analysis sequences having a molecular barcode that does not include the fixed base at the particular position. For example, in this embodiment, when the molecular barcode is clustered at Distance = 0 and also when the clustering is performed at Distance = 1 or more, the sequence having the molecular barcode that does not include the fixed base at the specific position is analyzed. It may further include excluding. In this case, the exclusion of the sequence having the molecular barcode not containing the fixed base at the specific position from the analysis may be performed before or after the clustering.
In each of the first and second embodiments, a sequence having a molecular barcode that does not include the fixed base at the specific position may be excluded from the sequence information of step (I), and is created in step (II). May be excluded from the analysis group or may be excluded from the analysis.
Alternatively, in each of the first, second, and third embodiments, in step (I) or after step (I), sequence information including a sequence containing the fixed base at the specific position is provided. You may get it. Alternatively, in the first embodiment, in step (II) or after step (II), a group consisting of a sequence containing the fixed base at the specific position may be obtained. That is, in the method for analyzing a nucleic acid according to the third embodiment, instead of the step (IIa), the step (IIb): in the step (I) or after the step (I), the identification is performed in the molecular barcode portion. Sequence information consisting of a sequence containing the fixed base at the position may be obtained; or Step (II): A sequence having a specific molecular barcode from the sequence information obtained in the above (I) or a sequence similar thereto. And selecting a sequence to form a group with the selected sequence, and step (IIc): in step (II) or after step (II) at the specific position in the molecular barcode portion. You may obtain the group which consists of the sequence containing the said fixed base. The sequence information or the group consisting of the sequence containing the fixed base at the specific position in the molecular barcode portion may be the sequence information containing the fixed base at all the specific positions. When the number of fixed bases is n (where n is a natural number), the sequence information or the group consisting of the sequence containing the fixed base at the specific position in the molecular barcode part is n, Alternatively, it may consist of a sequence containing mn fixed bases (where m can be 1, 2, 3, or a natural number ranging from 1 to n-1) at a specific position.

  In each of the first, second, and third embodiments described above, in step (III), a pair of an index and a molecular barcode other than the determined correct pair is determined as an index and a molecular barcode mispair. Or determined mispairs may be excluded from the analysis.

  In each of the above-mentioned first, second, and third embodiments, the method for analyzing a nucleic acid is based on the number of groups prepared based on a sequence having a specific molecular barcode or a sequence similar thereto. It may further comprise the step of determining the number of nucleic acid molecules of interest contained in the sample from which the molecules are derived.

  Those skilled in the art can understand that the first embodiment of the present invention, the second embodiment of the present invention, and the third embodiment of the present invention can be freely combined and implemented. For example, the first embodiment of the present invention can be combined with the second embodiment of the present invention, and the first embodiment of the present invention can be combined with the third embodiment of the present invention. . The first embodiment of the invention may be combined with the second and third embodiments of the invention. Furthermore, the second embodiment of the invention can be combined with the third embodiment of the invention.

Fourth Embodiment of the Present Invention A fourth embodiment of the present invention is a digital quantification method of a target nucleic acid molecule using a barcode sequence, which is the first embodiment, the second embodiment of the present invention, And a third embodiment, as well as implementations of embodiments selected from the group consisting of combinations thereof.

A fourth embodiment of the present invention is a digital quantification method of a nucleic acid molecule of interest using a barcode sequence,
(E) selecting a nucleic acid molecule containing the sequence of the target nucleic acid molecule from the obtained sequence information;
(F) clustering the nucleic acid molecules selected in (e) above for each unique molecular barcode sequence, and then identifying clusters having a plurality of sequences in the index nucleic acid molecule portion;
(G) In each of the clusters identified in (f) above, the pair of the index and the molecular barcode with the highest detection frequency is identified as the correctly indexed target nucleic acid molecule, and the other pairs of the index and the molecular barcode are identified. Deciding to be a mispair,
{May further include determining that the index is incorrect in Mispair},
Based on the number of unique molecular barcode sequence types linked to the correctly indexed target nucleic acid molecule (or the number of correctly indexed clusters of the target nucleic acid molecule), the target contained in the sample corresponding to that index Determine the number of nucleic acid molecules,
Can be a method. Here, in one embodiment, in step (g), the number of types of sequences of the unique molecular barcode linked to the correctly indexed target nucleic acid molecule (or the number of correctly indexed clusters of the target nucleic acid molecule) is determined. Alternatively, it may be determined as the number of target nucleic acid molecules contained in the sample corresponding to the index, and it is considered that the accuracy of quantification increases in principle as the number of reads increases.

The fourth embodiment of the present invention is
(A) separately obtaining a plurality of samples containing nucleic acid molecules (for example, DNA or RNA); {at least one of the samples contains a target nucleic acid molecule},
(B) A target nucleic acid molecule in which, before amplification of the nucleic acid molecule contained in the sample, an arbitrary molecular barcode is linked to each target nucleic acid molecule in each of the obtained plurality of samples, and different molecular barcodes are linked to each To obtain
(C) Prior to mixing a plurality of samples, a unique index is added to the target nucleic acid molecule for each sample containing a plurality of target nucleic acid molecules, and a library of target nucleic acid molecules in which different indexes are linked to each derived sample A step of obtaining (the order of Step B and Step C may be either first; and after Step (b) or (c), the nucleic acid molecule can be amplified to obtain an amplification product of the target nucleic acid molecule),
(D) In the mixture containing the amplification product of the nucleic acid molecule obtained after (b) and (c) above (the sample is mixed after the step (c), and the step (b ) May be carried out, and all the samples may be mixed after carrying out step (b), and the amplification product of the nucleic acid molecule linked with the molecular barcode is obtained after step (b), The sample may be mixed before obtaining the amplification product, and the sample containing the amplification product may be mixed after obtaining the amplification product), the index unique to each sample and the unique or arbitrary to each nucleic acid molecule of interest. Further comprising the step of sequencing the nucleic acid molecule to which the molecular barcode is added to identify the sequence of the index portion, the sequence of the molecular barcode portion and the sequence of the nucleic acid molecule portion linked thereto for each nucleic acid molecule. Good.

In the fourth embodiment of the present invention, for example, as described in the second embodiment, in (f), the clustering is
(I) In the sequence of the molecular barcode portion, a nucleic acid molecule group having the same sequence as the unique molecular barcode sequence is classified into the same cluster;
(Ii) In the sequence of the molecular barcode part, the nucleic acid molecule group having a sequence having a mismatch of up to 1 base with the sequence of the unique molecular barcode is classified into the same cluster;
(Iii) In the sequence of the molecular barcode portion, a nucleic acid molecule group having a unique sequence of the molecular barcode and a sequence having a mismatch of up to 2 bases is classified into the same cluster; or (iv) the molecular barcode It may be carried out by classifying nucleic acid molecule groups having a sequence having a mismatch of up to 3 bases with a sequence of a unique molecular barcode in a partial sequence into the same cluster.

In the fourth embodiment of the present invention, for example, as described in the second embodiment, in (e), the clustering is
Nucleic acid molecules having sequences sequenced with insertions or indels of bases (eg, up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of the molecular barcode portion are grouped into the same cluster. It may be performed by classifying. At this time, the molecular barcode having a fixed base described in the third embodiment may be used.

In the fourth embodiment of the present invention, for example, as described in the second embodiment, in (e), the clustering is
A nucleic acid molecule obtained by excluding a sequence sequenced as having an insertion or a deletion (indel) of bases (for example, up to 1 base, up to 2 bases, or up to 3 bases) in the sequence of a molecular barcode portion. It may be performed on groups. At this time, the molecular barcode having a fixed base described in the third embodiment may be used.

  By doing so, it is possible to correct possible nucleic acid sequence errors in the digital quantification method and improve the accuracy of digital quantification. That is, according to the present invention, by using a sufficiently large number of molecular barcodes as compared to the number of original nucleic acid molecules of interest in a sample, each nucleic acid molecule of interest is labeled by a molecular barcode having a different sequence. Accurate digital quantification is achieved by labeling and obtaining a sufficiently large number of reads relative to the original number of target nucleic acid molecules to detect all molecular barcodes attached to each target nucleic acid molecule. It will be possible.

Biology in the modern big data era requires accurate quantification of biomolecules in system-wide measurements. Because the quality of the analysis is highly dependent on the initial raw data. For this reason, digital quantification of nucleic acid molecules using DNA tags (referred to as “primer ID” ¹ , “UMI (unique molecular identifier)” or “molecular barcode”) has hitherto been performed. Being developed. This technique includes gene expression analysis by RNA sequence (RNA-Seq) ^2-7 , iCLIP (individual-nucleotide resolution UV cross-linking and immunoprecipitation) ⁸ , antibody repertoire analysis ⁹ , bacterial 16S rRNA gene analysis ¹⁰ , ¹¹ , and ChIP. -Nexus (chromatin immunoprecipitation experiments with nucleotide resolution through exonuclease, unique barcode and single ligation) ¹² has been used for many applications in next-generation sequencing platforms. These methods allow digitally accurate determination of the absolute number of molecules in a given sample, even in the presence of noise and / or bias in the measurement system. RNA-Seq using a molecular barcode, namely digital RNA-Seq (dRNA-Seq) ³ or quantitative RNA-Seq ¹³ is one of the most widely used applications of digital counting. dRNA-Seq works well even for small sample sizes and is often used for single cell gene expression analysis. Detection limits are important in such measurements. Because single cells have been shown to have many low copy RNAs ^13,14 , and the limit of detection indicates that there are many potentially undetected low copy RNAs, which are biological This can affect the subsequent interpretation of the phenomenon. Therefore, the investigation of the effectiveness of barcodes for absolute and digital quantitation is critical because the barcode system used determines the detection limit of nucleic acid quantification. Moreover, the simultaneous effectiveness of the bar code's ability to count high copy numbers is also important. Because, for example, random base barcodes have been used to label thousands of viral RNAs ¹ , and high-throughput single-cell RNA-Seq studies (where barcodes are used to label individual cells in a single sequencing run). Used to identify thousands of cells in ⁷⁾ .
The general procedure for digital quantification of nucleic acid molecules is as follows (see FIG. 1, panel A). (I) Uniquely tag each RNA (or complementary DNA or cDNA) or DNA with exogenously added DNA (molecular barcode) containing diverse sequences ^1-3 . (Ii) Amplify barcoded DNA or cDNA (generated from RNA if starting from RNA). (Iii) Sequence both the nucleic acid sequence of interest and the barcode sequence of the barcoded and amplified (c) DNA in tandem. (Iv) As has been theoretically proposed ¹⁵ , each nucleic acid of interest (or gene) to give the absolute copy number of the original nucleic acid of interest before amplification (ie, pre-amplification RNA or (c) DNA) ), The number of unique barcodes is quantified rather than the number of amplified molecules (the so-called “read number”). This scheme allows to exclude the effects of noise and / or bias generated at various steps during the measurement of the system (eg from amplification, sequencing, and / or analysis). To ensure that the digital counting system functions properly, each nucleic acid molecule of interest is guaranteed (or nearly guaranteed) to be uniquely tagged, and the measured number of unique molecular barcodes is Diverse barcode sequences must be used to equal the number of nucleic acid molecules of interest ^16,17 (first requirement below). In addition, it is empirically considered that a sufficient sequence depth is necessary for accurate counting ^18,19 (second requirement below).
Two types of barcode designs are typically used in digital counting schemes: sequence-limited barcodes (each barcode sequence is designed individually) and non-sequence-limited barcodes ("random"). Sometimes referred to as a "base" bar code). When sequence-restricted barcodes were previously used, the diversity of barcode sequences required for accurate quantification was estimated by theoretical calculations ¹⁶ , and for absolute quantification of barcoded molecules. The capacity of this technology has been experimentally confirmed ^3,16 . However, the use of sequence-limited barcodes has the following disadvantages: For measuring high dynamic range, many different individually designed barcode sequences must be prepared, which Not cost effective. Random (or pseudorandom) base barcodes have been used instead to increase cost while increasing the dynamic range of counts ^{2,4-9,11,12,18,20} . In this case as well, it should be determined that the sequence diversity of the barcode set is sufficient ^17,18 . However, unlike sequence-limited barcodes, only sequence changes in the barcode due to sequencing and / or amplification errors (newly generated barcode sequences from one of these errors can be false positives) ²¹ That is why this survey is not trivial. That is, the error can cause an overestimation of the number of molecules in the sample (in the case of sequence limited barcodes, all used barcode sequences are known, which also means that all unused barcode sequences are known. , Which means that the sequences resulting from the error can be identified and excluded). This problem is approached by using computational analysis to rule out errors based on the reasonable assumption that similar barcode sequences occur through errors originating from the same original barcode sequence. Furthermore, Sudbery et al. Recently demonstrated the efficacy of the random base UMI (molecular barcode) based on computer analysis by modeling errors for limited dynamic range (up to 100 molecules) ²² . However, the effectiveness of random base molecule barcodes for accurate digital counting is based on experiments that can clearly include effects that are not present in the theoretical model, especially with quantitative meaning ^7,20 and high dynamic range. Is not explicitly shown.
Here, we use random base molecule barcodes for digital quantification of the absolute number of barcoded DNA molecules when using a particular barcode design and after computer analysis. We will show experimentally what can be done. In order to investigate the effectiveness of the barcode itself by excluding other effects such as barcode addition and / or reverse transcription, which may vary in various applications, we have found that DNA molecules containing barcode sequences. Was synthesized and quantified by sequencing for amplified molecules (see panel A in Figure 1, dashed box). For accurate digital counting, we quantitatively investigated the above two requirements; (i) using a sufficiently large set of barcode sequences compared to the number of given molecules (see above). (As in panel B of FIG. 1), and (ii) sufficient sequence depth is achieved compared to the number of given molecules (panel C of FIG. 1). We then experimentally show that both the number of molecular inputs and the number of measured molecular outputs are consistent through a model measurement system that meets two requirements. In order to meet these two requirements, ie to ensure that the digital counting system works, we introduced a fixed base within the random barcode sequence for error detection and Perform barcode sequence clustering using the developed software, and use the information from the molecular barcodes to cross-contaminate between differently indexed samples and to identify the target nucleic acid sequence (template) in the mapping process. Misidentifications were identified and excluded. The present results show that accurate quantification of barcoded nucleic acid molecules in any given sample is high through proper barcode design (including minimal required barcode length) and sufficient sequence depth. We show that it can be achieved in the dynamic range (from 1 to more than 10 ⁴ and potentially up to 10 ¹⁵ molecules).
Hereinafter, in this example, the term “random” is used, but this term was randomly synthesized by an experimenter in order to secure a great variety of sequences without designing the sequences in this example. Means that.
[Method]
Library Preparation Single-stranded DNA templates containing random bases were purchased from Integrated DNA Technologies, Inc., Coralville, IA, USA (see Figure 13). The concentration of each template was 260 nm using a spectrophotometer (NanoDrop 1000; Thermo Fisher Scientific Inc., MA, USA) using the absorption coefficient described in the supplied specification sheet (Integrated DNA Technologies, Inc.). It was measured by absorption. Template DNA was stored at −30 ° C. at 50 μM in 0.1% (v / v) TWEEN20 (Sigma-Aldrich, St. Louis, MO, USA) solution. To control the concentration of DNA template for amplification, all templates were diluted with water (distilled water, deionized, sterilized, NIPPON GENE CO., LTD., Toyama, Japan) and 0.1% TWEEN20, and Mix in PCR tubes to final copy number. Amplification was carried out by PCR using MightyAmp (TAKARA BIO INC., Shiga, Japan) using 0.3 μM of each primer (see FIG. 14) in a 25 μL sample. Two tubes were prepared independently from 50 μM template stock and distinguished by the index designed into one of the primers (see Figure 14). Thermal cycling (ProFlex PCR system; Themo Fisher Scientific Inc.) was performed as follows: 98 ° C. half cycle; 98 ° C. 10 sec, 60 ° C. 10 sec, and 68 ° C. 1 min. 4 cycles; 98 ° C for 10 seconds, 60 ° C for 2 seconds, and 68 ° C for 1 minute 19 cycles; 68 ° C for 1/5 cycle; then incubated at 4 ° C. Then, the amplification product was subjected to column purification twice (DNA Clean & Concentrator ^™ -5; Zymo Research Corp, CA, USA), and the length distribution of the amplification product was analyzed by 2100 Bioanalyzer (Agilent Technologies, Inc., CA, USA). Confirmed using. Concentration was determined by qPCR kit (KK4602; KAPA Biosystems, Inc., MA, USA) using a real-time PCR system (7500; Themo Fisher Scientific Inc.).

Two samples with sequencing index (CGCTCATT: index A (index A), GAGATTCC: index B (index B)) 150 cycle kit v3 (Read 1: 100 cycles, Read 2: 50 cycles, Index 1: 8 Cycle) and MiSeq sequencer (Illumina, Inc.) in a single run. Read 2 was not used in the analysis as the sequence in Read 2 is part of the sequence in Read 1. The raw sequence data used for the analysis was deposited in GEO database GSE94895.

analysis
The Read 1 sequence was sorted by index A and B and a fastq file for each index was generated using MiSeq. In some cases, 100%, 32%, 10%, 3.2%, 1%, 0.32%, and 0.1% of reads were randomly sampled. The MiSeq fastq files were filtered by sequence length (≧ 34 bp and ≦ 39 bp long for short templates and ≧ 90 bp long for long templates). Alignment of the reads to the target nucleic acid sequence uses the target nucleic acid sequences of 11 types of templates as references (see “target” in FIG. 13), and uses Bowtie2 v.2.2.9 ²⁷ for long template (LT) and short template ( ST) was carried out individually. Essentially, uniquely mapped reads were used for subsequent analysis. The barcode region was 50 bp from the 5'end in the long template and 30 bp from the 5'end in the short template (see "barcode" in Figure 13) and these were extracted from the mapped reads. The fixed bases in the barcode region (up to 6 bases for short templates, up to 12 bases for long templates; see FIG. 13 “barcode”) were used for filtering and at least one fixed base mismatch was used. Bar codes with were excluded. The in-house software Nucleotide Sequence Clusterizer was then used to cluster the barcodes at Distance = 0, 1, 2, or 3. The number of clusters was considered to be the number of molecules before amplification. Reads with indexes A and B were integrated before clustering if index cross-contamination was considered. In the latter, multiple-mapped reads were also used for subsequent analysis. Then, after clustering, the minority leads were excluded if there were clusters containing multiple indexes. If the number of reads for index A and the number of reads for index B were the same, a coefficient of 0.5 was given to both index A and index B. Similarly, when misalignment was also considered, all reads mapped to the template with index A and index B were combined before clustering. When one read was mapped to multiple templates, a factor of 1 / (number of different templates) was given for each template. After clustering, minority reads were excluded if there were clusters containing reads and / or indexes mapped to multiple nucleic acids of interest. If the number of reads and / or indexes mapped to different templates is the same, a factor of 1 / (number of different templates and / or indexes) for each of the multiple mapped nucleic acids of interest and / or indexes. Was given. The number of leads in each process is as shown in FIG.

Nucleotide Sequence Clusterizer
For clustering, we coded in-house software named "Nucleotide Sequence Clusterizer" in C language. This tool performs clustering of DNA sequences using identified nucleotide positions in each sequence. This tool performs bounded single-link clustering: first each array is in its own cluster. If any two sequences differed from each other by no more than D mismatches, those clusters were integrated together. Here, D is a settable “Distance” parameter. This process continues until there are no more clusters to integrate, at which point the Nucleotide Sequence Clusterizer reports the number of clusters and the sequence within each cluster. Nucleotide Sequence Clusterizer is available on request.

  In this example, it was investigated whether or not the absolute number of DNA molecules contained in a sample could be accurately measured by a nucleic acid digital counting system using a random base barcode. As shown in FIG. 13, two major template DNAs were designed: six long templates (LT1-6) and five short templates (ST1-5).

As shown in FIG. 13, the nucleic acid molecules of LT1 to 6 are oriented from the 5 ′ end to the 3 ′ end,
It was designed to be the sequence of SEQ ID NO: 1-the barcode sequence-the target nucleic acid sequence-the sequence of SEQ ID NO: 2. The barcode sequences of LT1 to 6 and the nucleic acid sequences of interest are shown in SEQ ID NOs: 5 to 16.
Further, as shown in FIG. 13, the nucleic acid molecules of ST1 to 5 are directed from the 5 ′ end to the 3 ′ end side,
It was designed to be the sequence of SEQ ID NO: 3-the barcode sequence-the target nucleic acid sequence-the sequence of SEQ ID NO: 4. The barcode sequences of ST1 to 5 and the nucleic acid sequences of interest are shown in SEQ ID NOs: 17 to 26.

  All of these template DNAs contain a random base barcode shown as a molecular barcode group in panel A of FIG. 1, with the long template downstream of a 50 base barcode consisting of 38 random bases and 12 fixed bases. And a short template had an 8 nucleotide target nucleic acid sequence downstream of a 30 nucleotide barcode consisting of 24 random bases and 6 fixed bases (Fig. 13). reference). All templates also contained consensus sequences at both the 5'and 3'ends used for PCR amplification (see Figures 13 and 14). In the present embodiment, as model measurement samples, 40,000, 40,000, 4000, 300, 100, and 20 copies of LT1, LT2, LT3, LT4, LT5, and LT6, and 20000 copies of ST1 and ST2, and 4000 copies were prepared. Two identical samples were prepared, each containing ST3, ST4, and ST5. Amplify these templates in these two samples, distinguished by two different indexes (Index A and Index B) and sequence the amplified products using MiSeq to read 11,992,843 reads for Index A, Index B for On the other hand, 15,373,718 reads were obtained (see FIG. 15).

In this example, the sequences of indexes A and B were added to the template by including them in the reverse primer for amplification (see FIG. 14).
Sequence of reverse primer for amplification of index A (Rv primer in FIG. 14):
CAAGCAGAAGACGGCATACGAGAT AATGAGCG GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT (SEQ ID NO: 28)
Sequence of reverse primer for amplification of index B (Rv primer2 in FIG. 14):
CAAGCAGAAGACGGCATACGAGAT GGAATCTC GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT (SEQ ID NO: 29)
In the nucleic acid sequence of SEQ ID NO: 28, the underlined portion corresponds to the nucleic acid sequence of index A, and in the nucleic acid sequence of SEQ ID NO: 29, the underlined portion corresponds to the nucleic acid sequence of index B.

  Then, all reads are sorted by MiSeq for each index (A and B), the sorted reads are mapped to a reference consisting of the target nucleic acid sequence, and the number of reads read (that is, the number of amplified molecules). Instead of counting, the number of molecules for each index and template was quantified digitally by counting the number of unique barcodes (or the number of barcode clusters).

  Next, there are two requirements for accurate digital quantification in the presence of error (ie how many random bases are needed in the barcode to count a given number of molecules in a sample, and The number of leads (defined as "coverage") was examined (Figs. 2 and 8). Regarding the first requirement, by changing the number of random bases in each template on the computer (4-38 bases for LT and 4-24 bases for ST), each sorted index and The unique number of barcodes for the template was determined (Figure 2 panel A and Figure 8A; gray line). The number of unique barcodes determined increased dramatically with increasing number of random bases in the barcode. This suggests that a certain minimum number of random bases is needed to quantify a given number of molecules. Even if we artificially increase the number of possible barcode sequences (by increasing the length of the barcode), the number of original nucleic acid sequences of interest measured will not exceed the original copy number of 20,000. The plateau was expected to be at 20000 because it should not increase. However, the expected plateau in the region of higher random base numbers was not observed, and the number of unique barcodes determined increased monotonically as the number of random bases increased. With respect to the second requirement, the sequencing coverage was computationally modified by randomly excluding some of the reads, and the number of unique barcodes with the remaining reads for each index and template. Was determined (panel C of FIG. 2 and panel C of FIG. 8; gray line). If the digital counting scheme is working, once the coverage reaches a sufficient level, the number of unique barcodes identified should not depend on the coverage (sequence depth), so the plateau Will be observed in the plot. Increasing the sequence depth (ie, the number of times each barcode is read) should not increase the measured number of the original target nucleic acid sequence by more than 20000 of the original copy number, resulting in a plateau of 20000. I expected it to be. However, this expected plateau was not observed and the number of unique barcodes determined increased monotonically with increasing coverage. This suggests that under this condition the digital counting system needs improvement.

  The reason why no plateau was observed in these figures (panel A and panel C of FIG. 2 and panel A and panel C of FIG. 8) was finally sequenced compared to the input of the actual barcode sequence. In the barcode output, it can be explained by base changes, eg, substitution errors and insertion-deletion (indel) errors. To exclude substitution errors (possibly due to sequencing errors and / or polymerase amplification errors), the in-house software Nucleotide Sequence Clusterizer was used to cluster barcode sequences. In the clustering procedure, we introduced a parameter called "Distance": where Distance indicates the number of bases that differ between two given barcode sequences. For example, if one barcode sequence is exactly the same as another barcode sequence except for two base changes at any two positions, the Distance between these two barcode sequences is 2. Therefore, after clustering with Distance = 2 as a parameter, all barcode sequences in a given cluster are within Distance = 2 from at least one other barcode sequence in that cluster (included in the cluster Not every molecule is within Distance = 2 from all other sequences). In essence, it can be said that the original analysis for counting the number of unique barcodes without clustering was performed by clustering with Distance = 0. The number of barcode clusters was determined at Distance = 0, 1, 2, or 3 (see panel A of Figure 3 and panel A of Figure 9). It was expected that the number of barcode clusters would approach a certain value as the distance increased, if there were various types of barcodes for a given number of molecules, and that tendency was actually observed. In order to observe the effect of clustering on the two requirements for accurate digital quantification, the number of barcode clusters determined by using clustering with the longest Distance (Distance = 3) performed was used. It was plotted as a function of number (see blue line in FIG. 2 panel A and FIG. 8 panel A) and coverage (see FIG. 2 panel C and FIG. 8 panel C blue line). A more plateau-like curve was observed for both plots, but the number of barcode clusters determined still increased monotonically, especially as coverage increased.

  Next, it was attempted to exclude the effect of insertion-deletion (indel) type errors by excluding the reads containing mismatched bases at the positions of fixed bases in the barcode sequence of the sequenced reads (Fig. 13). reference). If the barcode sequence output contained mismatched bases at any of these fixed base positions, the deviation of the remaining bases from the designated "reading frame" defined by the fixed base positions was used. The resulting insertion or deletion of bases at another position in the barcode sequence was found. To investigate the effect of this process on the digital counting system, the position dependence of fixed bases in the barcode sequence was investigated. The number of barcode clusters when using one fixed base for this exclusion procedure was determined (see FIG. 3 panel B and FIG. 9 panel B). The number of barcode clusters decreased as the position of fixed base moved away from the sequence primer site. This is rational because the fixed base mismatch can detect indel type sequence changes that occur between the sequence start site and the fixed base position. In addition, the dependence of the number of fixed bases on the determined number of barcode clusters was analyzed. At this time, a fixed base located farthest from the sequence primer site was used (see FIG. 3 panel C and FIG. 9 panel C). When the number of fixed bases used was small, the number of determined barcode clusters decreased significantly, and as the number of fixed bases used increased, the number of determined barcode clusters became almost constant. To observe the effect of mismatch exclusion on the above two requirements for accurate digital quantification, the number of random bases (panel A of FIG. 2 and panel A of FIG. 8; green line) and coverage (FIG. 2). The number of barcode clusters determined was plotted as a function of panel C and panel C of FIG. 8 (green line). The mismatch exclusion process was performed using the highest number of fixed bases used (6 bases for short templates, 12 bases for long templates). As a result, a plateau-like curve was seen for all plots (panels A and C of FIG. 2 and panels A and C of FIG. 8), which indicates that indel error exclusion using fixed bases provided digital quantification. Indicates that it has been made more accurate.

As another problem, it was found that cross contamination between samples occurs. It is believed that this is causing a slight increase in the number of observed clusters in the plateau-like phase in the green line of panel C of Figure 2 and panel C of Figure 8. The two samples were each labeled with a different index (index A and index B) by amplification primers during PCR, and the two samples amplified by PCR were sequenced simultaneously in two separate tubes. When clustering barcodes using both index A and index B, a small fraction of barcode clusters containing both indexes were found. This has also been reported by Jaitin et al. ⁵ This could have occurred without cross-contamination as the barcoded templates for PCR amplification were randomly selected from the original template pool. However, even in the case of a short template, since there are theoretically 2.8 × 10 ¹⁴ (= ⁴²⁴ ) types of barcode sequences, the original template having completely the same barcode has two amplifications. It is considered that the possibility of being added to the tube is very small. Therefore, it is possible that either the PCR primer containing the specific index was mixed in the tube, the index sequence had an error, and / or index switching occurred in the sequencing process (Sinha. , R. et al., Biorxiv, 10.1101 / 125724 (2017)). To eliminate this effect, we first mixed all reads that were sorted into two indexes for each template, and clustered these mixed reads. If multiple indexes (two indexes in this case) were found in one barcode cluster, the barcode cluster was counted as having the index containing the highest number of reads in the sequence. Using this process, it was finally found that the determined number of clusters exhibited a plateau as a function of coverage (see the yellow line in panel C of Figure 2 and panel C of Figure 8). Importantly, in the blue lines of panel C of FIG. 2 and panel C of FIG. 8, it was observed that the number of clusters determined to increase the coverage ratio slightly increased, but the influence of index switching was observed. By the above process of exclusion, the cluster number showed a plateau with increasing coverage.

  Since it is unlikely that the same barcode sequence will be used for both indices, the sum of indices A and B was determined to confirm the first requirement for accurate digital quantification. The number of clusters was plotted (see yellow line in panel B of Figure 2 and panel B of Figure 8). The number of random bases used was within an acceptable range to perform an accurate digital quantification, as there was still a plateau.

  The above example shows that "index switching" can affect the accuracy of barcode clustering and the exclusion of index switching (misindex) that can occur when mixing and analyzing multiple samples. It is shown that the process improves accuracy and enables a digital quantitation system in which accuracy is not affected by coverage.

  Since cross-contamination between the samples was found, we next examined for misidentification in the process of mapping the lead to the reference. We followed a similar process we did for indexing issues. Here, all reads that were sorted into two indexes and mapped to either template were mixed, and then clustering was performed on the mixed reads. Then, when multiple templates and / or multiple indexes were found in one barcode cluster, the barcode cluster for the template and index that showed the highest number of reads among the sequenced reads was counted. However, throughout this process, no significant difference was observed in the number of determined clusters as a function of coverage (see red dotted line in Figure 2 panel C and Figure 8 panel C). This suggested that in this system misidentification does not occur very often. Since it is unlikely that the same barcode sequence will be used for both indexes and for all templates, we have confirmed that the first requirement for accurate digital quantification is index A, index B and all The number of determined clusters was plotted against the total template (see red dotted line in FIG. 2 panel B and FIG. 8 panel B). Since there was still a plateau, the number of random bases used was within an acceptable range for performing accurate digital counts, even when accounting for template misidentification. In this example, the effect of misidentification was less, but this process is important in assays where larger amounts of reference are used (eg RNA-Seq). This is because false identification can occur more frequently in such analyses.

  To further understand what is happening in the above analytical process, a histogram of coverage was created for each process (Figure 10). The histogram counting the number of unique barcodes (without any of the above treatments) had large peaks containing predominantly low read clusters. These low-read clusters artificially increase the output copy number of the target nucleic acid sequence when measured by this digital counting method (the artificial generation that is not present in the original sample due to sequence error, indel error, etc.). (Depending on the barcode sequence), this must be excluded for the system to make more accurate quantification (two requirements above). This peak was dramatically reduced after the first two processing steps, suggesting that the barcode sequences generated primarily by sequence errors were excluded by these processing steps.

  It was shown that the above barcode design and computer analysis fulfilled two of the above requirements for accurate digital counting when using four specific templates (ST1, ST2, LT1 and LT2) (FIG. 2 panels A-C and FIG. 8 panels A-C). The parameters were then optimized and these analyzes were applied to all templates containing a wide range of copy numbers from 20 to 40,000. Since the number of clusters determined when Distance = 2 as a parameter was already close to a constant value (see panel A of FIG. 3 and panel A of FIG. 9), Distance = 2 was used in the subsequent analysis. Regarding the number of fixed bases, when the number of fixed bases was 4, the number of determined clusters was close to a constant value (see panel C of FIG. 3 and panel C of FIG. 9). The number was set to 4 (for all templates, barcodes (FIG. 13) in which the 16th, 21st, 24th and 28th from the left are fixed bases were used). Cross-contamination of the index and misidentification of the template were also considered. All of the above quantitative analysis and insights could be used to accurately quantify nucleic acid molecules of interest using the digital counting scheme of the present invention. Based on these conditions, two requirements were examined for all molds to determine the dynamic range of this digital counting system (see panel A, panel B and FIG. 11 of FIG. 4). For the coverage dependency, 20 random bases were used for clustering (FIG. 4, panel A and FIG. 11), and for the dependency on the random base number, 10% of the original total read number was analyzed. It was decided to use (panel B of FIG. 4). Because it was thought that both parameters should still work based on previous initial analysis of the four original templates (FIG. 2, panel A, FIG. 2, panel C, FIG. 8 panel A and FIG. 8 panel C). When using less than 100% of the reads for analysis, the reads were randomly selected and the process was repeated 8 times to determine the mean and standard deviation (FIGS. 4A-4C and FIG. 11). As shown in Figures 4A-4C and Figure 11, there was a plateau for all templates in the plot as a function of number of random bases and coverage. This suggested that the parameters selected allowed accurate digital quantitation for a wide range of copy number templates.

  The determined number of barcodes in panel A of FIG. 4 and FIG. 11 was included in the sample tube before PCR amplification with a coverage of 12.6 to 20.9 (when 10% of the leads were sampled). It corresponded to the number of molds. These values were used to compare the number of molecular inputs determined by optical density with the number of molecular outputs determined by the digital counting method of the invention (see panel C of FIG. 4). As a result, a high correlation was shown between the number of input molecules and the number of output molecules (Pearson product moment correlation coefficient r = 0.990). This output / input ratio is in the range of 0.32-0.45 for the long template (LT) and 0.41-0.57 for the short template (ST) and may be explained by experimental error (eg PCR amplification). (At most) 7-step template dilution in preparation for (statistical error). This suggests that the parameter-based digital counting scheme presented in this example can quantify the absolute copy number of a nucleic acid molecule before PCR amplification.

Based on these results, the required number of random bases to count the absolute number of molecules in the presence of error can be presented (see Figure 5, panel A). The x-axis shows the number of inputs of the molecule to be measured and the y-axis shows the number of random bases when each curve in FIG. 4 panel B and FIG. 5 panel B reaches a relative cluster number of 0.95. Show. Panel B of FIG. 5 shows the dependence of relative cluster number on the number of random bases as done in panel B of FIG. 4, but for each template the misidentification exclusion process (cluster number Which did not have a significant effect). Include these data in panel A of FIG. 5 to show more data in the lower range for a given number of molecules, and, for example, to quantify about 10 ⁵ molecules with greater than 95% accuracy. Was found to require at least 16 random bases.

Experimentally, it was shown that at most 84,420 molecules (number of total LT input) were accurately quantified using 20 random bases (FIG. 4, panel B). This number is considered sufficient to count the number of RNA molecules for an individual gene in, for example, transcriptome analysis. In practice, the measurable number of molecules is limited by the capacity of the MiSeq sequencer.
Based on a simple linear regression to the experimentally measured data set, using a maximum of 38 random bases, with the number of random bases required (see Figure 12) depending on the number of given molecules. , About 10 ¹⁵ molecules can be quantified with the measurement system of the present invention. This dynamic range is far superior to the current capacity of commercial deep sequencers. As a result, the bottleneck of quantitative analysis with a wide dynamic range is no longer limited by the barcode design, but rather by the sequence throughput.

As described above, in the present Example, it was shown that the number of molecules in a given sample can be quantified by conducting a digital count using a hybrid type molecular barcode including a random base and a fixed base. . Here, an appropriate barcode design, sufficient sequence depth, and analysis method with appropriate parameters are used. This makes it possible to measure the number of nucleic acid molecules in a wide and high dynamic range and at low cost. Based on this result, one can suggest the number of random and fixed bases needed to count the number of a given barcode molecule in the presence of error (panel A of FIG. 5 and FIG. 12). This example also quantitatively demonstrated the additional functional advantage of molecular barcodes. That is, the molecular barcode is used to identify sample cross-contamination (caused by physical contamination of primers, errors in the index, and / or index switching in the sequencing process) and misidentification of the target nucleic acid sequence in the alignment process. did. In fact, as mentioned above, the former may solve a significant reported problem in next-generation sequencer platforms ^23,24 . Although the effects of errors may be dependent on library preparation and / or sequencing platforms, the effectiveness of random base barcodes has been demonstrated in common applications, and validation of barcode use presented here. The strategy for is applicable to various platforms. Furthermore, since the effectiveness of random base barcodes on barcoded molecules has been demonstrated, one of ordinary skill in the art, who can vary from application to application, can assess the effect or effectiveness of barcode addition. The present invention includes not only the gene expression analysis, iCLIP ⁸ , antibody repertoire analysis ⁹ , bacterial 16S rRNA gene analysis ¹⁰ , ¹¹ and molecular count in ChIP-nexus ¹² , but also cells ⁹ , ²⁵ , ²⁶ , virus ¹ , and barcode. It can be widely used for digital counting method of nucleic acid quantification using molecular barcode for other application uses. Recently, such as Single Cell Sequencing Solution (Illumina, Inc., CA, US and Bio-Rad Laboratories, Inc., CA, USA) and Chromium Single Cell 3'Solution (10x Genomics, Inc. CA, USA) are commercially available. The device may be used to make these applications. We believe that system biology will be promoted based on a large amount of experimentally obtained quantitative data.

Contents of Sequence Listing SEQ ID NO: 1: base sequence of 5'region of LT1 to 6 SEQ ID NO: 2: base sequence of 3'region of LT1 to 6 SEQ ID NO: 3: base sequence of 5'region of ST1 to 5 SEQ ID NO: 4: Base sequence of 3 ′ region of ST1 to 5 SEQ ID NO: 5: Barcode sequence of LT1 SEQ ID NO: 6: Target nucleic acid sequence of LT1 SEQ ID NO: 7: Barcode sequence of LT2 SEQ ID NO: 8: Target nucleic acid sequence of LT2 SEQ ID NO: 9: LT3 barcode sequence SEQ ID NO: 10: LT3 target nucleic acid sequence SEQ ID NO: 11: LT4 barcode sequence SEQ ID NO: 12: LT4 target nucleic acid sequence SEQ ID NO: 13: LT5 barcode sequence SEQ ID NO: 14: LT5 Target nucleic acid sequence SEQ ID NO: 15: LT6 barcode sequence SEQ ID NO: 16: LT6 target nucleic acid sequence SEQ ID NO: 17: ST1 barcode sequence SEQ ID NO: 18: ST1 target nucleic acid sequence SEQ ID NO: 19: ST2 barcode sequence SEQ ID NO: 20: ST2 objective nucleic acid sequence SEQ ID NO: 21: ST3 barcode sequence SEQ ID NO: 22: ST3 objective nucleic acid sequence SEQ ID NO: 23: ST4 barcode sequence SEQ ID NO: 24: ST4 objective Nucleic acid sequence SEQ ID NO: 25: barcode sequence of ST5 SEQ ID NO: 26: target nucleic acid sequence of ST5 SEQ ID NO: 27: sequence of forward primer for amplification SEQ ID NO: 28: sequence of reverse primer for amplification (for index A)
SEQ ID NO: 29: Sequence of reverse primer for amplification (for index B)

References

Claims (16)

A method for analyzing nucleic acids:
(I) subjecting a mixture of a plurality of target nucleic acid molecules having a molecular barcode and an index to sequencing to obtain sequence information;
(II) A sequence having a specific index or a sequence similar thereto and / or a sequence having a specific molecular barcode or a sequence similar thereto is selected from the sequence information obtained in (I) above, and selected. Creating a group with the arranged array,
(III) in the group created in (II) above, determining the pair of the index and the molecular barcode with the highest detection frequency as the correct pair of the index and the molecular barcode,
Including the method.   The method according to claim 1, wherein the nucleic acid molecule of interest to which at least the molecular barcode is added has been subjected to amplification before step (I).   A sequence similar to the sequence having the specific molecular barcode in step (II) is a sequence containing mismatch bases having a predetermined number of bases or less with the sequence having the specific molecular barcode in the molecular barcode sequence portion. The method according to Item 1 or 2.   The method according to any one of claims 1 to 3, wherein the molecular barcode has a fixed base at a specific position.   A sequence similar to the sequence having the specific molecular barcode in step (II) contains the fixed base at the specific position, and / or the position of the fixed base is shifted from the specific position. 5. The method of claim 4, selected on the basis of:   5. The method of claim 4, further comprising excluding a sequence having a molecular barcode that does not include the fixed base at the particular position from analysis. In the step (III), the pair of the index and the molecular barcode other than the determined correct pair is determined to be a pair of the index and the molecular barcode and excluded.
The method according to any one of claims 1 to 5.   The method further comprising the step of determining the number of the target nucleic acid molecule contained in the sample from which the target nucleic acid molecule is derived, based on the number of groups created by the sequence having a specific molecular barcode or a sequence similar thereto. The method according to any one of 1 to 7. A method for analyzing nucleic acids:
(I) subjecting a mixture of a plurality of nucleic acid molecules having a molecular barcode to sequencing to obtain sequence information;
(II) a step of selecting a sequence having a specific molecular barcode or a sequence similar thereto from the sequence information obtained in (I) above, and creating a group by the selected sequence;
Including the method.   A sequence similar to the sequence having the specific molecular barcode in step (II) is a sequence containing mismatch bases having a predetermined number of bases or less with the sequence having the specific molecular barcode in the molecular barcode sequence portion. Item 9. The method according to Item 9.   The method according to claim 9 or 10, wherein the molecular barcode has a fixed base at a specific position.   A sequence similar to the sequence having the specific molecular barcode in step (II) contains the fixed base at the specific position, and / or the position of the fixed base is shifted from the specific position. 12. The method of claim 11, selected based on:   12. The method of claim 11, further comprising excluding a sequence having a molecular barcode that does not include the fixed base at the particular position from analysis.   The method further comprising the step of determining the number of the target nucleic acid molecule contained in the sample from which the target nucleic acid molecule is derived, based on the number of groups created by the sequence having a specific molecular barcode or a sequence similar thereto. The method according to any one of 9 to 13.   The method according to any one of claims 9 to 14, wherein the nucleic acid molecule of interest to which at least the molecular barcode has been added has been subjected to amplification before step (I). A method for analyzing nucleic acids:
(I) subjecting a mixture of a plurality of nucleic acid molecules having a molecular barcode having a fixed base at a specific position to sequencing to obtain sequence information;
(IIa) excluding from the analysis a sequence having a molecular barcode that does not include the fixed base at the specific position;
(IIb) In step (I) or after step (I), a step of obtaining sequence information consisting of a sequence containing the fixed base at the specific position; or (IIc) the step (II) above (I). ) Further comprising the step of selecting a sequence having a specific molecular barcode or a sequence similar thereto from the sequence information obtained in step (4), and forming a group with the selected sequence, and in the step (II) or the step (II) After II), obtaining a group of sequences comprising the fixed base at the particular position;
Including the method.

Images