|
|
|
| e-Pub |
Previous | 
Home
Section: Research Program
Annotations
Word counting
Participants : Alain Denise, Daria Iakovishina, Yann Ponty, Mireille Régnier, Jean-Marc Steyaert.
We aim at enumerating or generating sequences or structures that are admissible in the sense that they are likely to possess some given biological property. Team members have a common expertise in enumeration and random generation of combinatorial structures. They have developped computational tools for probability distributions on combinatorial objects, using in particular generating functions and analytic combinatorics. Admissibility criteria can be mainly statistic; they can also rely on the optimisation of some biological parameter, such as an energy function.
The ability to distinguish a significant event from statistical noise is a crucial need in bioinformatics. In a first step, one defines a suitable probabilistic model (null model) that takes into account the relevant biological properties on the structures of interest. A second step is to develop accurate criteria for assessing (or not) their exceptionality. An event observed in biological sequences, is considered as exceptional, and therefore biologically significant, if the probability that it occurs is very small in the null model. Our approach to compute such a probability consists in an enumeration of good structures or combinatorial objects. Thirdly, it is necessary to design and implement efficient algorithms to compute these formulae or to generate random data sets. Typical examples that motivate research on words and motifs counting are Transcription Factor Binding Sites, TFBSs, consensus models of recoding events and some RNA structural motifs. When relevant motifs do not resort to regular languages, one may still take advantage of combinatorial properties to define functions whose study are amenable to our algebraic tools. One may cite secondary structures and recoding events.
Fast development of high throughput technologies has generated a new challenge for computational biology. The main bottlenecks in applications are the computational analysis of experimental data.
As a first example, numerous new assembling algorithms have recently appeared. Still, the comparison of the results arising from these different algorithms led to significant differences for a given genome assembly. Clearly, strong constraints from the underlying technologies, leading to different data (size, confidence,...) are one origin of the problems and a deeper interpretation is needed, in order to improve algorithms and confidence in the results. One objective is to develop a model of errors, including a statistical model, that takes into account the quality of data for the different technologies, and their volume. This is the subject of an international collaboration with V. Makeev's lab (IoGene, Moscow) and Magnome project-team. Second, Next Generation Sequencing open the way to the study of structuralvariants in the genome, as recently described in [44] . Defining a probabilistic model that takes into account main dependencies -such as the GC content- is a task o D. Iakovishina's thesis, in a starting collaboration with V. Boeva (Curie Institute).
Random generation
Participants : Alain Denise, Yann Ponty.
Analytical methods may fail when both sequential and structural constraints of sequences are to be modelled or, more generally, when molecular structures such as RNA structures have to be handled. The random generation of combinatorial objects is a natural, alternative, framework to assess the significance of observed phenomena. General and efficient techniques have been developed over the last decades to draw objects uniformly at random from an abstract specification. However, in the context of biological sequences and structures, the uniformity assumption becomes unrealistic, and one has to consider non-uniform distributions in order to derive relevant estimates. Typically, context-free grammars can handle certain kinds of long-range interactions such as base pairings in secondary RNA structures. Stochastic context-free grammars (SCFG's) have long been used to model both structural and statistical properties of genomic sequences, particularly for predicting the structure of sequences or for searching for motifs. They can also be used to generate random sequences. However, they do not allow the user to fix the length of these sequences. We developed algorithms for random structures generation that respect a given probability distribution on their components. Our approach is based on the concept of weighted combinatorial classes, in combination with the so-named recursive method for generating combinatorial structures. To that purpose, one first translates the (biological) structures into combinatorial classes, using the symbolic method, an algebraic framework developed by Flajolet et al. Adding weights to the atoms allows one to bias the probabilities towards the desired distribution. The main issue is to develop efficient algorithms for finding the suitable weights. An implementation was given in the GenRGenS software http://www.lri.fr/~genrgens/ , and a generic optimizer that automatically derives suitable parameters for a given grammar, is currently being developped.
In 2005, a new paradigm appeared in the ab initio secondary structure prediction [45] : instead of formulating the problem as a classic optimization, this new approach uses statistical sampling within the space of solutions. Besides giving better, more robust, results, it allows for a fruitful adaptation of tools and algorithms derived in a purely combinatorial setting. Indeed, we have done significant and original progress in this area recently [48] , [4] , including combinatorial models for structures with pseudoknots. Our aim is to combine this paradigm with a fragment based approach for decomposing structures, such as the cycle decomposition used within F. Major's group [47] .
Besides, our work on random generation is also applied in a different fields, namely software testing and model-checking, in a continuing collaboration with the Fortesse group at Lri [10] , [21] .
Programmed -1 ribosomal frameshifting
Participants : Patrick Amar, Jérôme Azé, Alain Denise, Christine Froidevaux, Yann Ponty, Cong Zeng.
During protein synthesis, the ribosome decodes the mRNA by assigning a specific amino acid to each codon or nucleotide triplet. Throughout this process the ribosome moves along the mRNA molecule three nucleotides at a time. However, encounters of specific signals found in mRNA from many viruses lead the ribosome to shift one nucleotide backward thus changing its reading frame. We aim at developing a new computational approach that is able to detect these specific signals in genomic databases in order to better understand the molecular choreography leading to the ribosomal frameshifting, which ultimately will help to rationally design new antiviral drugs. As candidates sequences are expected to be numerous, we aim at developing a ranking method to identify the most relevant sequences. Biological testing of these most promising identified candidates by our collaborators from IGM will help us to refine our computational method.
Knowledge extraction
Participants : Jérôme Azé, Jiuqiang Chen, Sarah Cohen-Boulakia, Christine Froidevaux.
Our main goal is to design semi-automatic methods for annotation. A possible approach is to focus on the way we could discover relevant motifs in order to make more precise links between function and motifs sequence. For instance, a commonly accepted hypothesis is that function depends on the order of the motifs present in a genomic sequence. Likewise we must be able to evaluate the quality of the annotation obtained. This necessitates giving an estimate of the reliability of the results. This may use combinatorial tools described above. It includes a rigorous statement of the validity domain of algorithms and knowledge of the results provenance. We are interested in provenance resulting from workflow management systems that are important in scientific applications for managing large-scale experiments and can be useful to calculate functional annotations. A given workflow may be executed many times, generating huge amounts of information about data produced and consumed. Given the growing availability of this information, there is an increasing interest in mining it to understand the difference in results produced by different executions.
Systems Biology
Participants : Patrick Amar, Sarah Cohen-Boulakia, Alain Denise, Christine Froidevaux, Loic Paulevé, Sabine Peres, Mireille Régnier, Jean-Marc Steyaert.
Systems Biology involves the systematic study of complex interactions in biological systems using an integrative approach. The goal is to find new emergent properties that may arise from the systemic view in order to understand the wide variety of processes that happen in a biological system. Systems Biology activity can be seen as a cycle composed of theory, computational modelling to propose a hypothesis about a biological process, experimental validation, and use of the experimental results to refine or invalidate the computational model (or even the whole theory).
Simulations and behavior analysis for metabolism modeling
A great number of methods have been proposed for the study of the behavior of large biological systems. Two methods have been developed and are in use in the team, depending on the specific problems under study : the first one is based on a discrete and direct simulation of the various interactions between the reactants, while the second one deals with an abstract representation by means of differential equations from which we extract various types of features pertaining to the system.
We investigate on the computational modelling step of the cycle by developing a computer simulation system, Hsim , that mimics the interactions of biomolecules in an environment modelling the membranes and compartments found in real cells. In collaboration with biologists from the Ammis lab. at Rouen we have used Hsim to show the properties of grouping the enzymes of the phosphotransferase system and the glycolytic pathway into metabolons in E. coli. In another collaboration with the SysDiag Lab (UMR 3145 ) at Montpellier, we participate at the CompuBioTic project. This is a Synthetic Biology project in the field of medical diagnosis: its goal is to design a small vesicle containing specific proteins and membrane receptors. These components are chosen in a way that their interactions can sense and report the presence in the environment of molecules involved in human pathologies. We used Hsim to help the design and to test qualitatively and quantitatively this "biological computer" before in vitro.
Given the set of biochemical reactions which describe a metabolic function (e.g. glycolysis, phospholipids' synthesis, etc.) we translate them into a set of o.d.e's whose general form is most often of the Michaelis-Menten type but whose coefficients are usuall very badly determined. The challenge is therefore to extract information as to the system's behavior while making reasonable asumptions on the ranges of values of the parameters. It is sometimes possible to prove mathematically the global stability, but it is also possible to establish it locally in large subdomains by means of simulations. We have developed a software Mpas (Metabolic Pathway Analyser Software) that renders the translation in terms of a systems of o.d.e's automatic; then the simulations are also made easy and almost automatic. Furthermore we have developed a method of systematic analysis of the systems in order to characterize those reactants which determine the possible behaviors: usually they are enzymes whose high or low concentrations force the activation of one of the possible branches of the metabolic pathways. A first set of situations has been validated with a research Inserm-Inra team based in Clermont-Ferrand. In particular we have been able to prove mathematically the decisive influence of the enzyme PEMT on the Choline/Ethylamine cycles (M. Behzadi's thesis, defended in 2011).
Comparison of Metabolic Networks
In the context of a national project, we study the interest of fungi for biomass transformation. Cellulose, hemicellulose and lignin are the main components of plant biomass. Their transformation represent a key energy challenges of the 21st century and should eventually allow the production of high value new compounds, such as wood or liquid biofuels (gas or bioethanol). Among the boring organisms, two groups of fungi differ in how they destroy the wood compounds. Analysing new fungi genomes can allow the discover of new species of high interest for bio-transformation.
For a better understanding of how the fungal enzymes facilitates degradation of plant biomass, we conduct a large-scale analysis of the metabolism of fungi. Machine learning approaches such like hierarchical rules prediction will be studied to find new enzymes allowing the transformation of biomass. The Kegg database contains pathways related to fungi and other species. By analysing these known pathways with rules mining approaches, we would be able to predict new enzymes activities.
Signalling networks
Amib and Inra-Bios (A. Poupon, Tours) are partners in a two years project Asam (2011-2012). This project aims to help the understanding of signalling pathways involving G protein-coupled receptors (GPCR) which are excellent targets in pharmacogenomics research. Large amounts of experiments are available in this context while globally interpreting all the experimental data remains a very challenging task for biologists. The aim of Asam is thus to provide means to semi-automatically construct signalling networks of GPCRS.