!1
Second International Conference in Code Biology
Jena, 16-20 June 2015
ABSTRACTS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Anna Aragno
Zachary Ardern
Marcello Barbieri
Gérard Battail
Andres Burgos
Han-liang Chang
Joachim De Beule
Peter Dittrich
Almo Farina
Elena Fimmel
Simone Giannerini
Louis Goldberg
Diego Gonzalez
Dennis Görlich
Markus Gumbel
Martin Hanczyc
Jannie Hofmeyr
Tim Ireland
Kristian Kraljic
Stefan Kühn
Ádám Kun
Barry McMullin
Chris Ottolenghi
Daniel Polani
Sonja Prohaska
Silke Sachse
Peter Stadler
Darko Stefanovic
Lutz Strüngmann
Eörs Szathmáry
Morten Tønnessen
Jochen Triesch
!2
Second International Conference in Code Biology – Jena, 16-20 June 2015
Dream Codes, II
Deciphering the Language of the deep Unconscious
Anna Aragno
After years garnering a reputation as outstanding research biologist, neuro-pathologist, medical
theorist of the etiology of hysteria, even Cocaine experimenter, Freud now daringly applies his
cumulative clinical observations and interpretive acumen to decoding the pictographic language
of the common dream.“The Interpretation of Dreams,” (1900) emerges amidst a combination of
wonder and ridicule from a Viennese medical community not uniformly swayed by Freud’s
psychical turn. Working in his legendary “splendid isolation” Freud has produced a masterwork
of scientific observation which, in chapters six and especially seven, presents a detailed analysis
of the two tiered structure; the motive force; the formal properties and compositional grammar
and syntax of a ‘primary process’ vocabulary, the language of the deepest unconscious in which
the meanings of dreams are spun. He has arrived at these landmark observations, and a
scientific method of interpretation, through the conceptual syntheses of the biogenetic law; the
unconscious as phylogenetically archaic; the primacy of early experience; the power of
repression, fixation, and regression, in sleep; and, most importantly, the sharply dichotomized
primary (impulsive) and secondary (inhibitory) processes, as cognitive principles of mental
functioning. Couched in his first topographical theory of mind, (Ucs. Pcs Cs), the cornerstone of
psychoanalytic metatheory, with the “Interpretation of Dreams” Freud established himself as the
fountainhead of a general dynamic psychology.
When listening clinically to verbal accounts of pictographic dream narratives we are hearing
the expression of sensory-affective experiences still tied to perceptual impressions and feelings
in the very process of being transformed into pre-cognized re-presentations. These iconic
clusters, stretching from sensory-kinetic-emotive physical experiences, to memories, motives,
fears, and desires; exhibiting proto-semiotic mechanisms of pre-linguistic tropes and cognition,
obey laws that reveal a “psychic reality” that is the true subjective experience of the dreamer.
With its metaphorical and manifest/latent structure, the dream is our MRI into the human
unconscious, as valuable an investigative instrument into primitive forms of early human
mentation as it is for uncovering individual psychical beliefs and states.
This presentation will cover in detail the mechanisms of the dream’s ‘primary process’
meaning-forms—the actual code of dreams— as laid out in chapter seven of the dream book
including an eye toward a contemporary re-interpretation of some Freudian ideas based on
updated integrations from an interdisciplinary palate of disciplines such as neuroscience,
semiotics, linguistics, dialogics, and narrative theory.
!3
Second International Conference in Code Biology – Jena, 16-20 June 2015
Deterministic and Optimizing Processes in Theories of
Genetic Code Origin
Zachary Ardern
School of Biological Sciences, University of Auckland
zard001@aucklanduni.ac.nz
In this paper I explore the relationships between chance, necessity and adaptation in different
theories of the origin of the genetic code. Two important areas of research in the origin of the
standard genetic code are stereochemical theories and work on various ways in which the codon
table is ‘optimised’.
I review the evidence for stereochemical theories, and the different kinds of optimality which
have been observed in the genetic code. Literature in the two fields seems to have largely
diverged, and there is a degree of tension between their basic premises – stereochemical
mechanisms are deterministic and act independently of any functional results, while optimality
implies some kind of adaptive process of optimisation. One finding which decreases the room for
compatibility between the two branches of theory is the fact that stereochemical binding
between codons or anticodons and their associated amino acids has predominantly been found
for complex amino acids generally thought to have been added to the code after its initial
stages. The concept of ‘coevolution’ of the code, constrained by the biochemical requirements
of amino acid synthesis, is another deterministic process which has been invoked to explain the
standard codon table. I also discuss how features of the code which are ‘adaptive’ in modern
biological systems such as eukaryotic cells might relate to early forms of life, arguing that
‘modern’ life has made remarkable use of features present from early stages of the code.
In light of these issues, I challenge the assumption that the genetic code was built up through a
pseudo-Darwinian gradualist mechanism involving searching vast areas of the information space
of possible codon tables. Instead I argue that while many questions are left unsolved, the early
development of the genetic code must have been constrained by many deterministic factors,
which in hindsight appear aligned with adaptations which biological systems subsequently
achieved.
!4
Second International Conference in Code Biology – Jena, 16-20 June 2015
Ancestral, Ancient and Modern Genetic Code
Marcello Barbieri
Dipartimento di Morfologia ed Embriologia
Via Fossato di Mortara 64a, 44121 Ferrara, Italy
brr@unife.it
The modern genetic code is a mapping between 64 codons carried by transfer-RNAs and 20
amino acids carried by 20 aminoacyl-tRNA-synthetases, each of which attaches one amino acid
to one or more tRNAs. The synthetases are specific proteins that can be made only by an
apparatus that already has a genetic code, and this gives us a classic chicken-and-egg paradox:
how could the genetic code come into existence if it is implemented by proteins that can be
produced only when the code already exists? The logical solution to this paradox is that the
modern apparatus of protein synthesis was preceded by an ancient apparatus where the amino
acids were attached to the transfer-RNAs not by proteins but by RNAs. The modern genetic code
based on protein-synthetases, in short, was preceded by an ancient genetic code based on RNAsynthetases. This ancient code, in turn, evolved from previous codes that here are collectively
referred to as ancestral genetic codes. The most primitive ribosomes were probably pieces of
ribosomal-RNAs that were sticking amino acids together at random and were producing
statistical proteins. The ancestral genetic codes, in other words, were ambiguous because a
codon could code for many amino acids, and a sequence of codons was translated some time
into a protein and some other time into a different protein. The modern genetic code, on the
other hand, is non-ambiguous because every codon codes for one and only one amino acid, and
evolved from an ancient code based on RNA-synthetases. This means that there have been three
distinct phases in the history of the genetic code: (1) the origin of the ancestral genetic code,
(2) the evolution from the ambiguous ancestral code to the ancient code, and (3) the evolution
from the ancient to the modern genetic code. Our problem is to figure out, at least in principle,
how could these three major transitions have taken place.
!5
Second International Conference in Code Biology – Jena, 16-20 June 2015
Genomic error-correcting codes shape the living world
Gérard Battail
Retired from E.N.S.T., Paris
email: gb@gerard.battail.name
Any material object is subjected to an erosion due to the chaotic behaviour of matter at the
molecular scale. The second law of thermodynamics states indeed that its physical entropy,
which measures how disordered it is, can but increase as time passes. Some sequence of symbols
written on such an object, however, can escape degradation by means of an error-correcting
code, a very sparse subset of all possible sequences to which the sequence to be protected must
belong. Its elements are referred to as words. The actually written codeword can unambiguously
be determined despite a number of symbol errors if the codewords are different enough, as a
result of the code scarcity (i.e., its redundancy) and of an even distribution of its words. In the
presence of a steadily increasing number of symbol errors, the sequence can be almost surely
conserved provided its recovery is attempted after a short enough time interval and the
recovered sequence is exactly rewritten, these operations being moreover regularly repeated.
Instead of an unavoidable erosion in the absence of coding, doing so ensures the sequence
conservation, except for a probability of failure which is very low if the code is efficient enough.
In case of failure, a sequence markedly different from the original one results. As another
codeword, it is conserved by the recovery process just as efficiently as the original one.
The DNA molecule is a material object at the molecular scale, so the permanence of the
symbolic sequence it bears, the genome, demands that it belongs to an error-correcting code.
That the living world is organized into discrete species, and not made of chimeras, directly
results and makes a taxonomy possible. This `main hypothesis', however, is not enough to
account for the better conservation of very old parts of the genome like the HOX genes. It
should be further assumed that the genomic error-correcting code is made of nested component
codes which successively appeared during the geological ages. This `subsidiary hypothesis'
explains that the taxonomy is actually hierarchical. The genome conservation process by almost
periodic recovery attempts (which are almost always successful) matches the existence of
successive generations in the living world, which should then be interpreted as regenerations.
The trend of evolution towards increasing complexity is another consequence of these
assumptions, since information theory tells that lengthening a code can improve its errorcorrecting ability. The Darwinian selection operating on the error-correcting ability of genomes
of increasing length then results in increasingly complex phenotypes.
The total number of extant and extinct species is very small as compared with the huge number
of nucleotide combinations that even the smallest genome can assume. The living world as a
whole thus matches the description given above of an error-correcting code: the genomes of
actual species are very sparse within the set of all possible nucleotide combinations. Besides
endowing any species with a unique label, a genome is also the recipe which controls the
assembly of a phenotype by means of some syntax which provides the needed redundancy. The
means by which living beings originate are thus identical to those which enable their
conservation.
!6
Second International Conference in Code Biology – Jena, 16-20 June 2015
Informational parasites in code evolution
Andres C. Burgos and Daniel Polani
a.c.burgos2@herts.ac.uk
daniel.polani@gmail.com
We consider the problem of the evolution of a code within a structured population of agents.
The agents try to maximise their information about their environment by “listening” to
information from other agents in the population. Which agents “listen” to which other agents is
determined by the structure of the population, which induces a “conversation graph” among
agents. The traditional use of information-theoretic methods would assume that every agent
knows how to “interpret” the informati-on offered by other agents. However, central to this is
the assumption that one “knows” which other agents one observes, and thus which code they
use. In our model, however, we specifically preclude that: it is not a priori clear which other
agents an agent is observing, and for understanding a code the agent either must know the
identity of the other agent with which it is communicating, or the code must be universally
interpretable. A universal code, however, introduces a vulnerability: a parasitic entity can take
advantage of it [1]. Here, we investigate this problem. We consider a parasite to be an agent
that tries to maximise its information about the environment, while, at the same time, inflicts
damage in the mutual understanding of the rest of the agents of the population. After
introducing a parasi-te in a structured population, we let the population respond to this new
addition by globally optimising the mutual under-standing of the agents via rearrangements of
their conversation graph. This optimisation from a global perspective indeed sel-ects the optimal
strategy, namely the isolation of the parasite. However, this assumes a global, unconstrained
update of the interactions between each of the agents. Now we consider, instead, a constrained
evolution of the population structure, in which agents choose their interactions indi-vidually,
according to their (limited) ability to identify who they are interacting with. In this case,
isolating the parasite is more difficult to achieve, since agents need to decide their interactions
based on local information only. The structure of a populati-on, as well as the code distribution
of the agents, both affect the individual capacity to identify other agents. We believe this
property to be crucial for the survival of populations, and a key issue for the development of an
immune response. Acknowledgements This research was supported by “H2020-641321 socSMCs
FET Proactive project”. The views expressed in this paper are those of the authors, and not
necessarily those of the consortium.
References [1] Csete, Marie & Doyle, John (2011). Bow ties, metabolism and disease. Trends in
Biotechnology, 22(9), 446–450.
!7
Second International Conference in Code Biology – Jena, 16-20 June 2015
Dual Coding of Memory in Classical Texts
Han-liang Chang
University Chair Professor, Fudan University, Shanghai, China
Professor Emeritus, National Taiwan University, Taipei, Taiwan
This paper discusses the theory of dual-coding and its application to information processing in
memory by rereading a few classical texts which deal, at least partially, with the literary
function of memory, such as Plato’s Timaeus and Symposium and Aristotle’s On Memory and
Recollection. Remote as they may seem, these texts demonstrate how information in memory
can be dually encoded, stored and retrieved as both visual and verbal representations, but once
when language’s meta-semiotic function is activated, all representation remains linguistic and
thus coding’s duality is compromised.
The Dual Coding Theory or Dual Code Theory (DCT), launched by Allan Paivio (1969, 1986,
2007), is a theory of memory which suggests that visual and verbal information act as two
distinctive systems and humans are capable of storing information in either visual codes and/or
verbal codes. The theory has been so popular in cognitive sciences that its usage has been
stretched from its original formulation and its application extended to other fields, such as life
science. For instance, Jesper Hoffmeyer and Claus Emmeche (1991) have proposed the notion of
“code-duality” to account for life – a notion that consists of two different linguistic “codes”, one
digital and the other analog. Compared with Paivio’s eclectic model, Hoffmeyer and Emmeche
have perhaps rightly banished the dubious visual coding and reinstated the primary, if solitary,
function of verbal language in representation. Since its inception nearly four decades ago,
Paivio’s assumption has been that “images and verbal processes are assumed to function as
alternative coding systems, or modes of symbolic representation” (1969: 243), and the two
interconnected processes perform independent functions in cooperative activity (2007). The
verbal and image systems are correlated, as one can think of the mental image of an automobile
and then describe it in words, or read words and then form a mental image. In brief, the DCT
identifies three types of processing: (1) representational, the direct activation of verbal or nonverbal representations, (2) referential, the activation of the verbal system by the nonverbal
system or vice versa, and (3) associative processing, the activation of representations within the
same verbal or nonverbal system. More recently, Paivio has refined the processing to include
five aspects: (1) verbal and nonverbal symbolic systems that cut across sensorimotor systems; (2)
the representational units of each system; (3) connections and activation processes within and
between systems; (4) organizational and transformational processes; and (5) conscious versus
unconscious processes (2007: 33).
Dual-coding in memory is quite common in Plato and Aristotle. Suffice it to give two
examples, one from each philosopher. In the prologue of the Timaeus, the speaker Critias
comments on his ability of retelling a childhood story he had overheard to this effect:
“Marvellous, indeed, is the way in which the lessons of one’s childhood ‘grip the mind,’ as the
saying is. For myself, I know not whether I could recall to mind all that I heard yesterday; but as
to the account I heard such a great time ago, I should be immensely surprised if a single detail
of it has escaped me…. [T]he story is stamped firmly on my mind like the encaustic designs of an
indelible painting.” (26C). The idea that memory trace appears as imagery is best illustrated by
Aristotle in his On Memory and Recollection. Where the faculty of recollecting is an ability, a
function of the soul, the so-called φαντάσια (phantasia or imagination), what is recollected are
φαντάσµατός (phantasmatos) (450a). Aristotle evokes two figures to “represent” what appears in
our recollection: τύπος (typos) a print type, and γραφὴ (graphē) which means
“writing” (450b16), both being doubly coded, first by “imagined” visual signs and then by
linguistic signs.
By discussing the phenomenon of dual-coding in the classical texts, the paper attempts to
bring into rapport the current discipline of Memory Studies with its ancient forgotten
predecessors.
!8
Second International Conference in Code Biology – Jena, 16-20 June 2015
Code Biology and the future of AI
Joachim De Beule
Data Scientist at Engagor, R&D department, Belgium
joachim@engagor.com
Artificial intelligence and cognitive science can be considered part of code biology at least to
the extent that they aim to unravel the secrets of human intelligence and that this will require
an understanding of the formation and workings of the neural codes. With the advance of big
data technologies such as deep learning, several distinguished scientists and entrepreneurs have
recently expressed their concern about the possibility of a super-AI surpassing human
intelligence and taking over control. At one level, I wish to temper these expectations and
refocus the debate by arguing that the new technology in fact does not provide solutions to
some fundamental problems such as the ‘symbol-grounding problem’, which is directly related
to the problem of the neural codes.
One of the main factors that made recent advances possible is the availability of huge amounts
of human-annotated training data, mainly due to the success of social media. For instance, each
time someone tags a friend in a picture posted on Facebook, data is produced that is used to
train face-recognition algorithms. As a result, the performance of these algorithms today
matches (and sometimes even surpasses) that of humans. But this would not have been possible
without the continuous stream of human-annotated training examples, which is why I argue that
the intelligence and codes being captured by the big-data crunching algorithms still essentially
are human intelligence and human codes. At the same time I wish to point to the possibility that
a novel intelligence is indeed emerging. The novel intelligence is not situated at the level of
neural codes however, but at that of cultural codes.
In this contribution, I will discuss this possibility and analyze recent disruptive advances from
the perspective of code biology, specifically from my earlier ‘sketch for a theory of evolution
based on coding’.
In short, social media as well as many other emerging Internet technologies today allow
individual people to extend into the social sphere and coordinate at a global level. This induces
conventionalization dynamics that can lead to the formation of novel cultural codes. Although
recent advances in AI do not in themselves lead to a novel, independent ‘super -intelligence’,
they do accelerate this process and ultimately may lead to a major or meta-system transition
giving birth to a genuinely novel intelligence or code-maker.
!9
Second International Conference in Code Biology – Jena, 16-20 June 2015
Discussion on
Code Hierarchies and Coarse Graining
Peter Dittrich
Bio Systems Analysis Group, Institute of Computer Science,
Friedrich-Schiller-Universität Jena, Germany
dittrich@minet.uni-jena.de
Joint discussion on: How do (organic) codes makeup hierarchies in living systems? How can we
precisely describe and investigate this hierarchy? How can we use this for coarse graining the
dynamics of living systems?
In the living world codes can be found at various scales, ranging from the molecular level over
neuronal systems to the social level. These codes are not only operating at different spatial but
also at different temporal scales. In this session we will discuss altogether whether the code
view provides a general approach for coarse graining the complex dynamics of living systems.
The genetic code indicates that this is at least partially possible, since the rules of the genetic
code can be interpreted as a coarse-graining of a complex dynamical molecular system involving
billions of molecular species.
Coarse graining from a mathematical point of view:
Given a dynamical process by a state transition function
f:X → X at a microscopic level with microscopic state
space X, i.e., x(t+1) = f(x(t)), a coarse graining is
associated with two mappings: one mapping h:X → Y
that maps microscopic states X to macroscopic states Y,
and one mapping g:Y → Y that describes the dynamics
on the macroscopic level, such that everything is
consistent, i.e., the mappings commute as h(f(x)) =
g(h(x)) for all microscopic states x from X. Note that Y is
usually much smaller than X.
How could this be related to a code view?
The coarse-graining function h describes which elements of the system (at the micro-state) are
considered to represent which sign, symbol, meaning, or code-element. For example, for the
genetic code, h maps (among others) chemical molecular species to words from {A,C,G,T}* .
The dynamics g on the macro-level includes the code rules, i.e. the mapping between the
“independent worlds”, which are both inside Y. For the genetic code, g includes the coding rules
mapping triplets to amino acids.
Procedure:
The plan is to present a small set of guiding questions like the following at the beginning of the
conference on Wednesday and ask for short contributions / answers to one or more of them to
be presented at the discussion session on Friday:
1. How do (organic) codes makeup hierarchies in living system
1. How can we precisely describe and investigate this hierarchy?
2. How can we use this for coarse graining the dynamics of living systems?
!10
Second International Conference in Code Biology – Jena, 16-20 June 2015
Acoustic codes and bird dawn chorus
Almo Farina
Department of Basic Sciences and Foundations
The University of Urbino, Campus Scientifico "Enrico Mattei"
61029 Urbino - Italy
almo.farina@uniurb.it
Dawn chorus is one of the most spectacular behaviors of birds, and one about which there has
been much speculation and the proposal of hypotheses without any definitive, univocal
explanations. At first sight the acoustic patterns that emerge from a dawn chorus seems a
chaotic blend of contemporarily songs from different species.
When analyzed in details
interesting patterns emerge that are the results of mechanisms operating at different temporal
and spatial scale under the influence of environmental proxies like the composition of bird
aggregations, the availability of resources and the vegetation structure of habitats.
A research conducted in five Mediterranean habitats during spring 2013 has found acoustic
patterns at the level of community that for their regularity could be considered like ecological
codes.
Dawn choruses start approximately one hour before the sunrise and stop suddenly at the
sunrise for few minutes, then a further song activity starts again and decreases after a couple of
hours.
Habitats rich in resources seem to have less differences between pre-sunrise and post-sunrise
song activity and some vegetation patterns (like evenness) seem to have a relevant influence.
!11
Second International Conference in Code Biology – Jena, 16-20 June 2015
Codon Usage in Circular and Comma-Free Codes
Elena Fimmel
Mannheim University of Applied Science
e.fimmel@hs-mannheim.de
Lutz Strüngmann
Mannheim University of Applied Science
l.struengmann@hs-mannheim.de
In 1957, Francis Crick suggested an ingenious solution for solving simultaneously the problem of
amino acid coding and frame maintenance. The idea was based on the notion of comma-free
codes. Even if the hypothesis of Crick revealed wrong, in 1996 Arquès and Michel discovered the
existence of a weaker version of such codes in eukaryote and prokaryote genomes: the so called
circular codes. Since then, circular code theory has provoked great interest and underwent a
rapid development.
In this talk the codon usage in maximal comma-free or self-complementary circular codes will be
discussed, i.e. we will investigate in how many of such codes a given codon participates. As the
main (and surprising) result it will be shown that the codons can be separated in very few classes
(three respectively six) with respect to their usage.
!12
Second International Conference in Code Biology – Jena, 16-20 June 2015
Are coding sequences optimized?
Simone Giannerini
University of Bologna, Department of Statistical Sciences
simone.giannerini@unibo.it
joint work with
Diego L. Gonzalez
CNR/IMM Bologna Section
and
Greta Goracci
University of Bologna, Department of Statistical Sciences
In this work we investigate the existence of universal optimizations in coding sequences. Usually,
these take the form of correlations between nucleotides that are observed in many organisms
and that might be related to error correction and energy optimization. We address the problem
motivated by a mathematical model of the genetic code introduced in Gonzalez(2008) and
further studied in Gonzalez et al.(2008, 2009), Giannerini et al.(2012). This new paradigm leads
to the definition of dichotomic classes that can be seen as nonlinear functions of the information
contained in a dinucleotide. Such classes, besides emerging naturally from the mathematical
model, represent precise biochemical interactions. We use the dichotomic classes as a binary
coding scheme for DNA sequences and study their mutual dependence by using suitable
resampling techniques. We find universal strong short-range correlations between certain
combinations of dichotomic classes. Remarkably, in some instances, we find that the sequences
of dichotomic classes are less correlated that purely random sequences and this suggests that
coding sequences can be seen as a minimum (or maximum) of some functional of the energy. The
results indicate that the paradigms of information/communication theory are essential for the
understanding of the organization of genetic information.
References
D. L. Gonzalez. The mathematical structure of the genetic code. In M. Barbieri and J.
Hoffmeyer, editors, The Codes of Life: The Rules of Macroevolution, volume 1 of Biosemiotics,
chapter 8, pages 111–152. Springer Netherlands, 2008.
D. L. Gonzalez, S. Giannerini, and R. Rosa. Strong short-range correlations and dichotomic codon
classes in coding DNA sequences. Physical Review E, 78(5):051918, 2008.
D. L. Gonzalez, S. Giannerini, and R. Rosa. The mathematical structure of the genetic code: a
tool for inquiring on the origin of life. Statistica, LXIX(3–4):143–157, 2009.
S. Giannerini, D. L. Gonzalez, and R. Rosa. DNA, frame synchronization and dichotomic classes: a
quasicrystal framework. Phylosophical Transactions of the Royal Society, Series A, Vol. 370,
Number 1969, 2987–3006, 2012.
!13
Second International Conference in Code Biology – Jena, 16-20 June 2015
The Role of Code Systems in Linguistic Communication
Louis J. Goldberg
State University of New York at Buffalo
goldberg@buffalo.edu
Inter-human communication enabled by both oral and written language is considered to be
coded. Nevertheless, the various disciplines engaged in the study of linguistic communication do
not fully comprehend the nature of biological coding systems and do not use the tenets of code
biology as a framework for viewing the results of their studies or for developing testable models.
This presentation is a beginning attempt to view language through the lens of code biology.
The basic code-elements of human oral linguistic communication are phonemes—the vowels
and consonants of a language. Phonemes are articulated by the speaker and conveyed as ‘digital’
packets of acoustic waves that traverse the atmospheric space between humans. Phonemes in
their manifestation as acoustic packets play the crucial role of conveying the acoustic signal
from one human to another, but they are only one of several phonemic code-elements. For
example, the problem in a hear-and-repeat paradigm of linguistic communication is to transform
an acoustic phonemic code-element articulated by a sender (e.g. any one of the consonants)
into the same consonant voluntarily repeated by a receiver. In essence, the consonant
articulated by the sender enters the environment, goes through the brain of the receiver, is
spoken by the receiver and comes back into the environment. The receiver-detected acoustic
pattern has a particular structural signature. It is obvious that the medium of the body and brain
cannot support travel of an acoustic signal. The information encoded in the pattern of the
acoustic signal undergoes many transformations as it travels through the receiver’s auditory
receptor system, to the cerebral cortex and then to the vocal musculature. Remarkably, the
information contained in the pattern of the acoustic signal is conserved in the code-element
transfer process through body and brain. The brain has transformed itself, along with its
ancillary sensory receptor and musculoskeletal systems, into a code generating and receiving
entity to enhance the effectiveness of inter-human communication.
The invention of writing introduced an additional set of code-elements into linguistic
communication. These are the graphemes, the orthographic representations of phonemes. The
manner in which graphemes as code-elements are transferred throughout the body and brain are
similar to that described above for phonemes. Writing brought literacy into human social
systems. Literacy is considered to be a uniquely human cross-modal cognitive process. Auditory
phonological representations in brain (the phonemes of speech) become associated in brain with
orthographic representations (the graphemes of reading and writing). To use the vague
conceptual terminology of cognitive science and neurobiology, this requires the association of
the processing stream of phonological information with the visual-object processing stream.
The exceptional cross-modal abilities of human language are currently of high interest to
cognitive scientists. Consider that literate humans can hear-to-write (take dictation); read-towrite (copy); read-to-speak (give speech from text). Consider also that humans can accomplish
these operations without external input, e.g., think-to-write, think-to-speak and think-to-think
(self-to-self conversations). The code biology perspective leads to the proposition that these
kinds of code-based, cross-modality operations are accomplished through the actions of an
adaptor system in brain association cortex that governs the obvious rule-based coded
correspondences among writing, reading and speaking. Although the words, code, encode and
decode, are casually used in many publications concerning linguistic communication, an
understanding of the coding processes as elaborated by code biology is not in evidence. In my
view, the cognitive science and neurobiological approach to linguistic communication would be
greatly enriched by an appreciation of the principles of code biology.
!14
Second International Conference in Code Biology – Jena, 16-20 June 2015
On the origin of degeneracy in the genetic code
Diego L. Gonzalez
gonzalez@bo.imm.cnr.it
The genetic code is the first molecular code identified in living organisms. Its very existence
poses many fundamental questions about the origin and evolution of life. Here we present an
analysis from a new perspective on this challenging problem. For this analysis we rely on both, a
biological hypothesis for the origin of degeneracy, and a mathematical model describing exactly
the degeneracy distribution of different variants of the genetic code. The biological hypothesis
is based on primeval symmetry properties of nucleic acids and of molecular adaptors [1]. The
mathematical model, instead, is based on redundant integer number representations systems, in
particular, on binary, non-power, positional representation systems [2]. First, we analyse the
degeneracy structure of the vertebrate mitochondrial genetic code [1]. Because this is the
simplest variant among all known versions of the genetic code, different authors proposed it as a
paradigmatic model for the early genetic code, that is, the code that preceded the universal
genetic code characteristic of the Last Common Universal Ancestor (LUCA). Our tessera code,
based on the above mentioned symmetry properties, is the natural candidate as the ancestor of
the early code, that is, it is a parsimonious solution for the pre-early code. Second, we describe
different steps that predated the appearance of the pre-early code until the so-called
archetypical code [3,4], and beyond, near the origin of amino acid coding. Such steps strongly
suggest that, as in many different physical examples, the evolution of the genetic code
proceeded through successive symmetry breaks. These correspond to a progressive ambiguity
reduction, that is, progressive decreasing degeneracy in amino acid coding. Finally, going
forward in time, we can interpret within the same framework different schemes associated to
post-transcriptional modifications in modern codes that putatively appeared at least after the
establishment of the early code. From the viewpoint of code theory, we propose that the
present forms of the genetic code are profoundly determined by physico-chemical constraints,
mainly the symmetry properties of the molecules involved in primeval protein synthesis.
Moreover, we observe a strong conservation of the degeneracy distribution that can be followed
from the pre-early code to the present vertebrate mitochondrial genetic code and include the
preservation of other relevant symmetry properties. Because the proposed solution for the preearly code exhibit intrinsic features of error detection/correction systems, we hypothesize that
the evolutionary pressure for decoding efficiency represents the main cause for its selection.
However, we cannot exclude that other important biological functions might be associated to
the conservation of the degeneracy properties of the genetic code.
References
[1] Gonzalez Diego L., Giannerini S., and Rosa R., "On the origin of the mitochondrial genetic
code: Towards a unified mathematical framework for the management of genetic
information", Nature Precedings, http://dx.doi.org/10.1038/npre.2012.7136.1, (2012).
[2] Gonzalez Diego L., Giannerini S., and Rosa R., "On the origin of degeneracy in the genetic
code", preprint (2015).
[3] Gonzalez Diego L., "The Mathematical Structure of the Genetic Code", in: The Codes of Life
(book chapter 6), Marcello Barbieri Editor, Springer Verlag (2008).
[4] Watanabe, K. and Yokobori, S., "How the early genetic code was established? -inference from
the analysis of extant animal mitochondrial decoding systems-", in Chemical Biology of
Nucleic Acids: Fundamentals and Clinical Applications, RNA technologies, Erdmann, V.,
Markiewicz, W. and Barciszewski, J., eds., chap. 2, 25–40 Springer Verlag, (2014).
!15
Second International Conference in Code Biology – Jena, 16-20 June 2015
The relational basis of molecular codes – Towards
hierarchies of molecular codes
Dennis Görlich
Institute of Biostatistics and Clinical Research,
Westfälische Wilhelms-Universität Münster, Germany
Dennis.goerlich@ukmuenster.de
The existence of codes in biology, such as the genetic code, is a fact that enables us to search
for other codes in biological systems. Some other cellular subsystems have been proposed to
implement an encoded relationship between molecular “worlds”, e.g. the histone code. We may
study and identify each code individually, i.e. without its relation to other codes, but these
relations may exist and can be described formally. Recently, I proposed the idea of the relational
basis of molecular codes (Görlich, 2014) which formalizes the relations between molecular codes
mathematically based on the formalization presented in (Görlich & Dittrich, 2013). Two types of
code relations can be identified (i) a linkage of codes, illustrating the situation that codes are
executes subsequently as part of an information processing cascade (ii) a nesting of codes. Both
types enhance a systems code capacity, i.e. the number of mappings that could be realized.
Code linkage can be present in two subtypes (a) where the meanings of a first codes, can be
used as signs in a second code, and (b) a setting where the meanings of a first code have an
effect on the context of the second code, i.e. the first code controls the coding rules of the
second code. I will discuss these situations, sketch proofs for the mathematical properties of the
relations, and propose the concept of hierarchies of codes, or code networks.
Görlich D (2014) The relational basis of molecular codes. Biosemiotics 7:249-257.
Görlich D, Dittrich P (2013) Molecular Codes in Biological and Chemical Reaction Networks. PLoS
ONE 8(1): e54694.
!16
Second International Conference in Code Biology – Jena, 16-20 June 2015
BDA-generated Models of the Genetic Code
in current and ancient genetic codes
M. Gumbel1, E. Fimmel1, A. Danielli2, L. Strüngmann1
1M.
Gumbel (m.gumbel@hs-mannheim.de), E. Fimmel (e.fimmel@hs-mannheim.de),
L. Strüngmann (l.struengmann@hs-mannheim.de)
Faculty of Computer Science, Mannheim University of Applied Sciences,
Paul-Wittsack-Str. 10, D-68163 Mannheim, Germany
2A. Danielli (alberto.danielli@unibo.it)
University of Bologna, Department of Pharmacy and BioTechnology, Via Irnerio 42, 40126 Bologna, Italy
In this talk we introduce the concept of models of the genetic code which are based on binary
dichotomic algorithms (BDAs) [1][2]. A BDA divided the set of 64 codons into two disjoint classes
of size 32 each and provides a generalization of known partitions like the Rumer dichotomy [3]
or the Parity class dichotomy [4]. We analyze what partitions can be generated when a set of
different BDAs is applied sequentially to the set of codons. An algorithm is developed that scans
for interesting BDA-generated models. The search revealed that those models are able to
generate code tables with very different number of classes ranging from 2 to 64. It is analyzed
whether there are models that map the codons to their amino acids. A perfect matching is not
possible. However, there are models that describe the standard genetic code with only little
errors. There are also models that map all 64 codons uniquely to 64 classes showing that BDAs
can be used to identify codons precisely. This could serve as a basis for further mathematical
analysis using e. g. coding theory. The hypothesis that BDAs might play a role at the decoding
center of the ribosome is discussed and first results on BDAs in the context of ancient genetic
codes are presented.
REFERENCES
1.
2.
3.
4.
M. Gumbel, E. Fimmel, A. Danielli, L. Strüngmann: On Models of the Genetic Code generated by
Binary Dichotomic Algorithms. Biosystems 128 (2015) 9-18.
E. Fimmel, A. Danielli, L. Strüngmann: On dichotomic classes and bijections of the genetic code, J.
Theor. Biology, Volume 336, Pages 221-230 (November 2013).
Y.B. Rumer: Systematization of codons in the genetic code, Doklady Akademii nauk SSSR 1969 Aug;
187(4):937-8.
D.L. Gonzalez: The Mathematical Structure of the Genetic Code, In Barbieri (Ed.): Codes of Life:
The Rules of Microevolution, Springer, 2008
!17
Second International Conference in Code Biology – Jena, 16-20 June 2015
Synthesis of living systems :
perspectives from artificial life
Martin M Hanczyc
University of Trento
Centre for Integrative Biology (CIBIO)
Trento, Italy
Martin.Hanczyc@unitn.it
My work is focused on understanding the fundamental principles of living and evolving systems
through experimental science. To this end I build synthetic systems where dynamic life-like
properties emerge when self-assembled systems are pushed away from equilibrium. I will
present an experimental model of bottom-up synthetic biology: chemically-active oil droplets.
This system has the ability to sense, metabolize and the potential to evolve. Specifically, I will
present how sensory-motor coupling can produce chemotactic motile droplets. In addition I will
present how we add information to our supramolecular structures in the form of ssDNA that then
governs the assembly and disassembly of higher order architectures.
!18
Second International Conference in Code Biology – Jena, 16-20 June 2015
On the logical necessity for a manufacturing code
in self-fabricating systems
Jan-Hendrik Hofmeyr
Centre for Studies in Complexity, University of Stellenbosch,
Stellenbosch 7600 South Africa
and
Wissenschaftskolleg zu Berlin, Berlin 14193, Germany
jhsh@sun.ac.za
We, the living, all do something astonishing that no non-living thing can do. Unlike, say, a car,
which needs you or a mechanic to replace or repair its parts, we can fabricate all our parts by
ourselves. Self-fabrication underlies all those properties so often used to define life: growth,
metabolism, reproduction, development, adaptation, even evolution. Without explaining selffabrication we can explain neither life itself, nor its origin. In traditional fabrication, as carried
out by humans, there is an external agent, the artisan. In Aristotelian terms, such an agent is
the efficient cause. In self-fabrication, however, there is no external agent; all the efficient
causes are produced within the system. Such a system is, therefore, closed to efficient
causation. But such closure requires an efficient cause to be produced by another efficient
cause, which in turn must be produced by another efficient cause, which seems to lead to an
infinitely regressing chain of efficient causes. I shall present a formal model for self-fabrication
that resolves this conundrum. It models the fundamental and universal biochemical process of
polymerization as the creation of symbol strings in a formal language by production rules, which
are themselves created as symbol strings from within the language in order to ensure closure to
efficient causation. But, just as a linear polymer must fold into a specific three-dimensional
structure before it can become functional, so a symbol string that describes a production rule
must in some way acquire semantic meaning before it can do its job. This implies that just as
the correct folding of a polymer requires the appropriate chemical context, so a production rule
string can only acquire meaning in an appropriate linguistic context. Through a series of logical
steps I build on this formal scaffolding to construct a complete model of a self-fabricating formal
language. I show that avoiding the infinite regress of rules that produce rules that produce rules,
and so on, necessitates the separation of the description of the sequence of symbols from the
construction of the sequence; this in turn necessitates the prior encoding of the sequence and
the subsequent translation of the encoded form of the sequence by a specialised set of rules
that are also produced within the system. Such a coding system has been called a manufacturing
code. The beauty of this linguistic model of self-fabrication is not only that it maps perfectly
onto cell biochemistry as we know it, but that it demonstrates that the problem of closure to
efficient causation in linear polymer systems, the key to self-fabrication of all living things, can
logically be solved in only one way.
!19
Second International Conference in Code Biology – Jena, 16-20 June 2015
Encoding Space
Tim Ireland
Senior Lecturer, Programme Leader MA Architectural Design
Leicester School of Architecture
De Montfort University Leicester, UK
tireland@dmu.ac.uk
Space like many concepts (such as those of information and mind) escapes concrete definition.
Taking a biological stance on the matter space is perceived a dynamic emergent manifold which,
like a dangling carrot hanging before us but beyond our reach, evades comprehension. As such,
trying to define what space is may be deemed to be a waste of … space. However, space is a
fundamental property of life (Hall 1966). Biotic beings are spatial and exist in a predominantly
spatial manner, affecting and affected by those things they share the world with; and although
one being may be deemed of no significance to another it holds potential significance. The
spider’s web, for example, works because flies cannot detect its presence. The web exists even
though the fly doesn’t perceive it; until, that is, it is too late for the fly.
As a relationship which is produced supra-subjectively across agents, space has objective
properties. However it is not an object per se, but a ‘form’, as Kant claimed of ‘sense
perception’ because it is tangible in the sense that it can be perceived – and can be acted upon
as a sign. At this most primal level it is a pattern of interaction. As an artefact, formed through
an organism’s capacity to affect and manipulate its environment, space is objectified. The
definitive manifestation of physical space is thus an artefact, which embodies the spatiality of
the organism that created it. Having progressed from congregating around fire, humankind
constructs buildings serving purposes beyond basic physiological needs, such as cultural and
personal expression. These patterns are established through codes-of-being, propagated in a
dialectical process of interaction between organism and environment, effecting artefacts such as
nests, structures and buildings.
The French Marxist philosopher and sociologist Henri Lefebvre recognised the mental and
physical aspects of space as intertwined and mediated through what he called spatial-practice;
referring to habitual tendencies cast into the artefacts and structures we inhabit. In other
words, lived-space is the assimilation of physical and mental space. His spatial-code expresses a
tri-dialectic process whereby space is created and creative effected through the relations
between subjects, their space and surroundings, meaning space is social because it unfolds
through interaction. As a social product, he states, “[s]pace is neither a ‘subject’ nor an ‘object’
but rather a social reality – that is to say, a set of relations and forms” (Lefebvre 1995: 116).
Emphasising the social dimension of being in the world Lefebvre stresses interaction is both
mental and physical. “Space is social morphology: it is to lived experience what form itself is to
the living organism, and just as intimately bound up with function and structure” (Lefebvre
1995: 94). His model provides a basis for scrutinising spatial practice but doesn’t provide a
means for, as the genetic code does, generation, because it does not define rules which may be
interpreted and built upon. So if space is a product of mediating mental and physical aspects of
being what is the underlying code that enables this mapping? Is there an underlying elemental
code on which codes-of-being are founded? If so would it facilitate designing, and thus enhance
the built environment, and how might this capacity be applied?
!20
Second International Conference in Code Biology – Jena, 16-20 June 2015
Genetic Code Tool (GenTool)
Elena Fimmel, Lutz Strüngmann and
Kristian Kraljic
Institute of Applied Mathematics
Faculty for Computer Sciences Mannheim
University of Applied Sciences
68163 Mannheim, Germany
k.kraljic@kooperationen.hs-mannheim.de
The Genetic Code Tool (GenTool) is a tailor-made software tool, to support and simplify the
creation and mathematical analysis of sets and sequences of genetic code tuples, such as
trinucleotide codes and their
corresponding amino acids. The main purpose of the tool is to provide a simple and modern user
interface and a tool-aided approach, to create, transform and visualize sequences or sets of
tuples, as well as to perform tests and analysis on those. Written in Java the tool is easy to
enhance via an open interface, to add an arbitrary number of custom transformations, splits,
tests and analyses, as well as new controls to input and visualize genetic code sequences and
sets.
!21
Second International Conference in Code Biology – Jena, 16-20 June 2015
Catch and release:
attachment and dispersal codes in a bacterial biofilm
Stefan Kühn and Ruenda Loots
Dept. of Biochemistry, University of Stellenbosch,
Stellenbosch, 7600, South Africa
stefankuhn21@gmail.com
Biofilms are naturally occurring, surface-bound aggregates of microbes. Bacteria, Archea, fungi,
algae, and protozoans are able to accumulate within an autogenic three-dimensional matrix of
extracellular polymeric substances. Such aggregates have been independently called ‘microbial
cities’, ‘secret societies’, and even the ‘bacterial twitter’. The need for communication both
inter- and intra-species is paramount if such densely populated, highly organized ‘societies’ are
to survive and thrive. However, before our humble microbes blossom into an entire city, the first
brick must be laid. This necessitates a mechanism by which a microbe may sense its
environment, relay the signal, and translate it into action. We propose that this mechanism,
despite experimental difficulties, exists and can be explained from the perspective of code
biology. The system wherein the microbe – P. aeruginosa – senses its environment and translates
that signal, via an intricate array of biomolecules, to states of attachment, growth, or dispersal
are the attachment and dispersal codes. To aid us in the explanation of these codes we draw
upon the body of knowledge provided by Barbieri and his work on organic codes. Since the term
`code’ has seen a number of uses – not all of which conform to the precepts of an organic code
as provided by code biology – we shall
endeavour to test the precepts of the bacterial communication code against those established by
code biology. Drawing on our current knowledge of P . aeruginosa biofilms we will narrate the
attachment, development, and dispersal of just such a biofilm from within the context of an
organic code. Further, we will show that the environment, signalling molecule, and effect
threesome of such bacterial attachment and dispersal systems conform to the sign–adapter–
meaning
trichotomy as laid out in code biology. A particular change in the microbe’s
environment is sensed and acts as an organic sign which is then relayed by a specialized signaling
molecule that acts as an adapter. This adapter molecule then translates the sign into attachment
or dispersal – the biological meaning.
!22
Second International Conference in Code Biology – Jena, 16-20 June 2015
The Coding Coenzyme Handle Hypothesis revisited
Ádám Kun
kunadam@elte.hu
According to the Coding Coenzyme Handle Hypothesis, the genetic code first served to reliably
attach amino-acids to ribozy-mes, so the amino-acids can offer their chemical versatility to the
ribozymes. The amino-acids are attached to small oligonucle-otides (handles), which in turn can
attach to ribozymes via hydrogen-bonds. Complementary triplets forming kissing-hairpins can
offer a strong enough binding. If amino-acid were coenzymes fostering catalysis, then (1) there
are not many sites on a ribozyme onto which the handles could attach, and (2) we find mostly
codons of amino-acids with known catalytic importance at these sites. We investigate these
questions by analysing the secondary structures of random RNAmolecules and counting the
number of attachment sites as well as their identity. It turns out that the codons for lysine,
asparagine, glutamine, isoleucine, leucine and phenylalanine are frequently found in singlestranded regions of RNSs, thus offering potential attachment sites to the coenzyme-handles.
According to some theories, only the middle nucleotide or the first two nucleotides were used in
coding in the past. If we ac-cept that only the middle nucleotide is important, then codons with
A and U in the middle are the ones that could be good handles. We find many of the catalytically
important amino-acids (histidine, aspartic acid) in the A column. This is good for the theory!
However, we find apolar amino-acids in the U column. There is some debate in the literature
about the early role of amino-acids, and catalysis is only one of them. They could have been
employed in a structural role, helping ribozyme to fold properly, as well as helping membrane
functions (RNA cannot submerge into membranes). Our results suggest that according to the
extant code coupled with the Coding Coenzyme Handle Hypothesis, they could serve both roles.
!23
Second International Conference in Code Biology – Jena, 16-20 June 2015
Computational modelling of the evolution of symbolization
Barry McMullin
Dublin City University, Ireland
barry.mcmullin@dcu.ie
The paradigmatic example of a biological code is the relation between the nucleotide sequence
in mRNA and amino acid sequence in a corresponding protein. It is of the essence of this code
that its physical realisation is reflexive: the key protein enzymes responsible for the specificity
of the code (the t-aminoacylsyntheseses) must themselves be self-consistently coded in
corresponding genes in order for the coding system to be sustained (and ultimately, to be
reproduced). I will present a schematic, but fully functional, computer model (or, more
precisely, “virtual world”) which attempts to capture this key characteristic of an arbitrary, but
self-referentially bootstrapped and sustained, coding system. This is based on realisation of the
generic architecture for machine self-reproduction first formulated and articulated by John von
Neumann in the late 1940’s. While the model is simple, it is still conceptually powerful, and
allows practical investigation of real evolutionary phenomena perturbing the initially defined
coding system in complex ways. Some concrete examples of these phenomena will be presented.
!24
Second International Conference in Code Biology – Jena, 16-20 June 2015
Futile metabolic cycles and memory generation
Chris Ottolenghi
Service de Biochimie Métabolomique et Protéomique
Hôpital Necker Enfants Malades
149 rue de Sèvres, 75015 Paris, France
chris.ottolenghi@parisdescartes.fr
In individual organisms, metabolic cycles mediate a range of functions such as substrate
transport across biological boundaries (e.g., the malate-aspartate shuttle); energy storage (e.g.,
creatine phosphate); heat production (e.g., mitochondrial uncoupling proteins); or complex
reactions involving ‘catalytic’ metabolites (e.g., acetyl-CoA combustion via the Krebs cycle).
Metabolic cycles are termed “futile” when they do not show any net effect except (supposedly)
wasting energy. Indeed, futile cycling has been implicated in several mechanisms of pathology.
But what for the apparently futile cycles that occur under physiological conditions? I will review
examples focusing on distinct lipogenesis-lipolysis cycles. Available data will be discussed in
relation to the concepts of cell memory generation, stemness, and transdifferentiation.
!25
Second International Conference in Code Biology – Jena, 16-20 June 2015
Collectives, Codes and Conventions
Daniel Polani
Department of Computer Science
University of Hertfordshire
Hatfield AL10 9AB, United Kingdom
daniel.polani@gmail.com
We discuss the role of information in self-organization of collectives, information-limited
acquisition of information about the world and how agents can improve the quality of their
(limited and incomplete) world description by "agreeing" on conventions to describe it.
It turns out that collectives indeed can capture more structural properties about the
environment than the individual if information exchange is limited. In other words, informationlimited codes may sharpen agents' perception of the structure of their environment.
!26
Second International Conference in Code Biology – Jena, 16-20 June 2015
I/O Maps and Codes in Regulatory Systems
Sonja Prohaska
Computational EvoDevo
Institute of Computer Science
University of Leipzig
sonja@bioinf.uni-leipzig.de
Regulation is inherently dynamic and always involved with the transmission of information from
inputs to outputs. To keep inputs and outputs from mixing, they need to be of different types.
Consequently, an mapping of values for the input variable onto corresponding output values
can be formally represented as a typed I/O map. This map encapsulates a causal relation and
may or may not constitute a code.
In this contribution I will discuss the potential relevance of the computer science concepts
"indirection" and "arbitrariness" for biological codes.
Furthermore, I will elaborate on the importance of organizational or logical structures in the
input values and the mapping. To complete the list of conceptual elements of regulatory
systems, I shall mention "feedback" and "memory" and use the well studied regulatory system,
the lac operon, as an example.
!27
Second International Conference in Code Biology – Jena, 16-20 June 2015
Olfactory Coding - from Odor Molecules to Brain Activity
Silke Sachse
Olfactory Coding Group
Max Planck Institute for Chemical Ecology
Department of Evolutionary Neuroethology
Hans-Knoell-Strasse 8
D-07745 Jena
Germany
ssachse@ice.mpg.de
We are investigating how odors are coded and processed in the Drosophila brain to lead to a
specific odor perception. The basic layout of the first olfactory processing centers, the
vertebrate olfactory bulb and the insect antennal lobe, is remarkably similar. Odors are encoded
by activated glomeruli in a combinatorial manner. Drosophila melanogaster provides an
attractive model organism for studying olfaction, as it allows genetic, molecular and
physiological analyses. The talk will summarize our recent insights into coding strategies of
ecologically relevant odors yielded by morphological and functional analysis of the neuronal
populations present in the antennal lobe and aim to find the link to odor-guided behavior.
!28
Second International Conference in Code Biology – Jena, 16-20 June 2015
Codes in RNA – How to design RNA Structures
Peter F. Stadler
Bioinformatik, Inst.f.Informatik, Univ.Leipzig
Haertelstrasse 16-18, D-04107 Leipzig, Germany
peter.stadler@bioinf.uni-leipzig.de
The sequences of RNA molecules implicitly determines the molecules' structure. Largely, this
encoding is determined by the thermodynamics of RNA folding, to a certain extent also by the
kinetics of the folding process and by protein that interact with the RNA.
Meanwhile we understand the rules that link sequence and structure of RNAs sufficiently well to
use them for the rational design of RNAs with intricate functions: riboswitches are abundant
regulators of both transcription and translation in procaryotes that rely on environmentally
triggered structural changes.
After an introduction in the simple rules of the RNA structure "code", the presentation will focus
on the inverse folding problem and practical design strategies for functional RNAs.
!29
Second International Conference in Code Biology – Jena, 16-20 June 2015
Implementing Molecular Logic Gates, Circuits, and Cascades
Using DNAzymes
Darko Stefanovic
darko@tau.cs.unm.edu
The development of electronic digital logic was one of the greatest technological achievements
of the 21st century, and exponential increases in the computational power of commerciallyavailable microprocessors meant that electronic computers are now ubiquitous and
indispensable in the modern world.
Contemporaneous advances in molecular biology made it clear that information processing is a
fundamental capability of all biological systems.
Subsequent rapid progress in that field
progressed in parallel with the development of consumer electronics, and elucidated many of
the mechanisms behind
biological information processing.
Given that the information processing capabilities of
biological systems were evolved over millions of years, it is fascinating to consider whether we
can construct synthetic molecular computers that may be more compact and more robust than
their natural or electronic counterparts. When we say that molecules compute, what we usually
mean is that an assembly of molecules detects certain inputs, typically the presence or absence
of other molecules, and responds by producing one or more output signals, which may take the
form of the release of an output molecule or the generation of a detectable fluorescent signal.
The goal of a molecular computer scientist is to engineer the intervening molecular system so
that the pattern of output signals is related to the pattern of input signals by the desired logic
function. From an unconventional computing perspective, the development of molecular
computers offers intriguing possibilities to implement extremely low-power computation and to
implement autonomous computational systems that can survive and thrive in environments
hostile or inaccessible to silicon microprocessors, such as within the bloodstream or within living
cells. The compact nature of DNA has been previously exploited to demonstrate high-density
information storage, but in our context the fact that billions of molecules exist in each
experimental system may make it feasible to execute massively parallel computations in a very
small volume, or to implement novel computational architectures that compute using the
dynamics of interactions between molecular circuit components.
Our experimental work focuses on catalytic nucleic acid chemistry, in particular, DNAzymes
(deoxyribozymes), which are DNA-based enzymes that can cleave or combine other nucleic acid
strands. DNAzymes are now known to occur in nature, and the known DNAzyme catalytic motifs
have been isolated in in vitro evolution experiments. We turned DNAzymes into logic gates by
augmenting them with up to three input-binding modules that regulated the catalytic activity of
the DNAzyme based on the pattern of input strands observed in the solution. The cleavage
reaction catalyzed by the DNAzyme served as the reporting channel, and we exploited the
combinatorial chemistry of DNA to enable us to build systems that processed multiple signals
simultaneously in a single solution, with the different information streams tagged by different
DNA sequences. Thus, each DNAzyme unit implemented a logic gate with up to three inputs, and
we constructed a set of such gates complete for Boolean logic. I will review our designs for
DNAzyme-based molecular computers, their integration in large-scale parallel gate arrays
exhibiting sophisticated logical and temporal behaviors, and our recent attempts to diversify
into sequential logic cascades. Our early approach to molecular computing included the first
reported complete set of nucleic acid-based logic gates, which we then used to produce
autonomous molecular computing systems that implement well-known logic circuits such as
adders and large-scale game-playing automata. I will then discuss current work in which we
have extended this approach to achieve signal propagation in DNAzyme signaling cascades and
have begun to apply these new techniques to biodetection applications.
!30
Second International Conference in Code Biology – Jena, 16-20 June 2015
Dinucleotide comma-free codes and their connection to
dinucleotide circular codes
Lutz Strüngmann
Mannheim University of Applied Science
l.struengmann@hs-mannheim.de
Elena Fimmel
Mannheim University of Applied Science
e.fimmel@hs-mannheim.de
The presence of circular codes in mRNA coding sequences is postulated to be involved in
informational mechanisms aimed at detecting and maintaining the normal reading frame during
protein synthesis.
Most of the recent research is focused on trinucleotide circular codes. However, also
dinucleotide circular codes are important since dinucleotides are ubiquitous in genomes and
associated to important biological functions. In this talk a group theoretic approach for
investigating maximal dinucleotide comma-free codes will be shown and symmetry properties of
such codes will be highlighted. Moreover, we describe a construction principle for such codes,
their close connection to maximal dinucleotide circular codes and
provide a graph representation that allows to visualise them geometrically. The results obtained
can be interpreted as a weaker formulation of the well known hypothesis by Crick about frameshift-detecting codes without commas.
!31
Second International Conference in Code Biology – Jena, 16-20 June 2015
Towards major evolutionary transitions theory 2.0
Eörs Szathmáry
Center for the Conceptual Foundations of Science, Parmenides Foundation,
Kirchplatz 1, D-82049, Munich, Germany;
MTA-ELTE Theoretical Biology and Evolutionary Ecology Research Group,
1c Pázmány Péter, H-1117 Budapest, Hungary
szathmary.eors@gmail.com
The impressive body of work on the major evolutionary transitions in the last twenty years calls
for a reconstruction of the theory, although a two-dimensional account (evolution of
informational systems and transitions in individuality) remains. Significant advances include the
concept of fraternal and egalitarian transitions (lower level units like and unlike, respectively).
Multilevel selection first without, then with the collectives in focus is an important explanatory
mechanism. Transitions are decomposed into phases of origin, maintenance and transformation
(i.e. further evolution) of the higher-level units, that helps reduce the number of transitions in
the revised list by two so that it is less top-heavy. After the transition units show strong
cooperation and very limited realized conflict. The origins of cells, the emergence of the genetic
code and translation, evolution of the eukaryotic cell, multicellularity, and the origin of human
groups with language are reconsidered in some detail in the light of new data and
considerations. Arguments are given why sex is not in the revised list as a separate transition.
Some of the transitions can be recursive (e.g., plastids, multicellularity) or limited (that share
the usual features of major transitions without a massive phylogenetic impact, such as the
micro-and macronuclei in ciliates). During transitions new units of reproduction emerge, and
establishment of such units requires high fidelity of reproduction (as opposed to mere
replication).
!32
Second International Conference in Code Biology – Jena, 16-20 June 2015
Umwelt codes
exemplified by Umwelt alignment in corvids
Morten Tønnessen,
University of Stavanger
mortentoennessen@gmail.com
I have previously suggested that there are different Umwelt codes, which can be categorised as
either CODEfix (fixed Umwelt codes) or CODEflex (flexible Umwelt codes). While neural codes are
examples of CODEfix, ecological codes are instances of CODEflex. Some corvids, including crows,
evidently prosper in part due to their relationships with human settlements and anthropogenic
food sources. However, actual human-corvid relationships are typically somewhat distanced –
likely because, for one thing, corvids are often treated by humans as pest species. In this paper
I will look into the Umwelt alignment (cf. Tønnessen 2014) between corvids including crows on
one hand and human beings on the other. My hypothesis is that Umwelt alignment must typically
involve ecological codes (Umwelt codes).
What is Umwelt alignment? In dictionaries, “alignment” signifies an adjustment to a line, or
arrangement into a straight line; a state of agreement or cooperation; the proper positioning or
state of adjustment of parts in relation to each other; etc. Crucially, alignment can denote
processes or states of fitting-in with others. If we define Umwelt alignment as the process of
adjustment by one creature to the presence and manifestation of other Umwelt creatures (and
further, to abiotic Umwelt objects and meaning factors), we realize that every Umwelt dweller
on this planet conducts Umwelt alignment on a regular basis, as manifested over time in
concrete functional cycles (see Uexküll 2010 [1934/1940]: 49).
Not all Umwelt alignment is mutual and cooperative. If Umwelt alignment is a universal
phenomenon, then there must also be Umwelt alignment among natural enemies, and among
competitors. Keeping a certain distance can be seen as emblematic of Umwelt alignment. The
spatial distribution of specimens is central in human and animal social life as well as in terms of
human ecology and general ecology, and is relevant in the current context to the extent that
spatial distribution is arranged by way of Umwelt creatures with deliberate adjustment to the
presence of others. Here we observe not only various forms of natural (autonomous) Umwelt
alignment, but further instances of coerced Umwelt alignment, a phenomenon which is enforced
and motivated more or less exclusively by human utility.
The regulatory mechanism of Umwelt alignment thus ranges from symbiotic strategies to more
competitive forms of coexistence. In all cases, however, various forms of synchronicity are key. A
study of Umwelt alignment between corvids and humans can expectedly shed light on humancorvid co-evolution, corvid Umwelten, and current human ecology.
References
Tønnessen, Morten 2014. Umwelt Trajectories. Semiotica 198 (Special Issue on zoosemiotics,
guest-edited by Timo Maran): 159–180.
Uexküll, Jakob von 2010 [1934/1940]. A foray into the worlds of animals and humans – with A
theory of meaning (Posthumanities 12), Joseph D. O’Neil (trans.). Minneapolis & London:
University of Minnesota Press.
Acknowledgement: This work has been supported by the research project “Animals in Changing
Environments: Cultural Mediation and Semiotic Analysis” (EEA Norway Grants/Norway Financial
Mechanism 2009–2014 under project contract no. EMP151).
!33
Second International Conference in Code Biology – Jena, 16-20 June 2015
Information Coding in the Brain
Jochen Triesch
Frankfurt Institute for Advanced Studies
Ruth-Moufang-Straße 1,
60438 Frankfurt am Main, Germany
triesch@fias.uni-frankfurt.de
How does the brain encode information, i.e., what is the neural code? And how does the brain
acquire the code(s) that it uses? Although these questions have always been a central topic of
Neuroscience research, our current understanding is still very limited. I will first give an
overview of the coding problem from the perspective of Neuroscience. Then I will focus on the
efficient coding hypothesis, which claims that animal brains exploit redundancies in their
sensory inputs to encode information more efficiently. Finally, I will discuss a recent extension
called active efficient coding, which generalizes the theory to so-called active perception,
which includes movements of the sense organs.