CatMOD

Contents

• 1   Description
• 2   Getting Started
  • 2.1   Requirements
  • 2.2   Installation
  • 2.3   Running the tests
• 3   Usage
  • 3.1   Preprocess
  • 3.2   Data processing
  • 3.3   Extracting features
  • 3.4   Predicting
• 4   Support
• 5   Contributing
• 6   Citation
• 7   License
• 8   Changelog

1   Description

CatMOD is a CatBoost-based feature-concatenation model for RNA modification recognition from ONT DRS.

2   Getting Started

2.1   Requirements

CatMOD Project is a python3 package. To use CatMOD, python version 3.9 or higher is required.

• python >= 3.9
• catboost
• h5py
• numpy
• pysam
• rich
• scipy

2.2   Installation

git clone https://github.com/CatMOD/CatMOD.git
cd CatMOD
conda create -n catmod -y python=3.9
python setup.py install

or

git clone https://github.com/CatMOD/CatMOD.git
cd CatMOD
conda env create -f catmod.yml

or

git clone https://github.com/CatMOD/CatMOD.git
conda create -n catmod -y python=3.9 catboost h5py numpy pysam rich scipy
conda activate catmod

2.3   Running the tests

3   Usage

3.1   Preprocess

Using Guppy and Tombo processes ONT fast5 files.

guppy_basecaller --input_path $fast5_folder --recursive --fast5_out --save_path $guppy_folder --flowcell $FLOWCELL --kit $KIT --num_callers $THREADS
multi_to_single_fast5 --input_path $guppy_folder --save_path $single_folder --threads $THREADS --recursive
tombo resquiggle --rna --processes $threads --overwrite --fit-global-scale --include-event-stdev $single_folder $REFERENCE

3.2   Data processing

catmod data_process

3.3   Extracting features

catmod extract_features --bed $sample_bed --ref $REFERENCE --align $ont_bam --current $ont_current --threads $THREADS --output $datasets_folder

3.4   Predicting

catmod predict --bed $sample_bed --datasets $datasets_folder --model /path/to/CatMOD/models/wheat_pretrained.cbc.cbm --threads $THREADS --output $datasets_folder

4   Support

For any bugs/issues, please feel free to leave a message at Github issues. We will try our best to deal with all issues as soon as possible.

For any suggestions/comments, please send emails to: Shang Xie xieshang0608@gmail.com, Jingjing Zhai zhaijingjing603@gmail.com.

5   Contributing

6   Citation

7   License

8   Changelog