Menu:

Research projects

The Laboratoire de Bioinformatique des Génomes et des Réseaux (formerly Service de Conformation des Macromolécules Biologiques et de Bioinformatique) has a long-standing experience in bioinformatics and in the large scale analysis of genomes, transcriptomes and proteomes. The main research lines focus on the following topics.

The methods developed in the lab are made available to the scientific community via web-based interfaces and as distributed software.

Analysis of regulatory sequences

Non-coding sequences play an essential role in living organisms, by controlling the regulation of gene transcription. This regulation is mediated by interactions between transcription factors and short DNA motifs present at specific locations in the genome. The computational prediction of transcription factor binding sites can provide valuable hypotheses about gene regulation. The availability of an increasing number of fully sequenced genomes allows us to combine the detection of regulatory signals with comparative genomics, in order to study the conservation and divergence of cis-regulatory elements, and, thereby, decipher the evolution of genetic regulation.

Since 1997, Jacques van Helden developed methods to predict regulatory signals in non-coding sequences. These approaches are based on the detection of statistically significant motifs in genomic sequences. The approach has been thoroughly tested with microbial genomes (yeast, bacteria), and is currently extended to metacellular organisms (insects, vertebrate).

The Regulatory Sequence Analysis Tools (http://rsat.bigre.ulb.ac.be/rsat/) are available to the academic community via a Web server and as Web services.

Analysis of mobile genetic elements (ACLAME)

The ACLAME project (http://aclame.ulb.ac.be) is driven by Raphaël Leplae, in collaboration with Ariane Toussaint and Gipsi Lima-Mendez. A database dedicated to the collection and classification of mobile genetic elements (MGEs) from various sources, comprising all known phage genomes, plasmids and transposons is provided to the scientific community. The project aims at building a comprehensive classification of the functional modules of MGEs at the protein, gene, and higher levels. Tools for MGEs analysis and detection in bacterial genomes are also available on the web site.

Analysis of biomolecular networks

The function and evolution of living organisms is based on networks integrating various types of molecular interactions: transcriptional regulation, protein interactions, metabolic reactions, signal transduction, ...

A large bunch of information is available about the different pieces of such networks. During the second half of the 20th century, biochemists and molecular geneticists have characterize such molecular interactions on a case-by-case mode. More recently, high-throughput methods (DNA microarrays, TAP-TAG, ChIP-chip, ...) have been designed to unraveal thousands of interactions in a single experiment. These data sets can be combined to obtain a global network synthetizing our current perception of molecular interaction networks. Interpreting such networks is a challenging goal for modern biologists.

Our laboratory develops dedicated bioinformatics and statistical approaches for the interpretation of molecular interaction networks: assessing the reliability of individual interactions, extracting functional modules from the global network, deciphering the relationshp between network topology and emergent behaviour, ...

Our Network Analysis Tools (Neat, http://rsat.bigre.ulb.ac.be/neat/) are available to the academic community as a Web server and as Web services.

Analysis of metabolic pathways

Most of our current knowledge about metabolism is based on a handful of model organisms. This knowledge represents a tiny fraction of the wide variety of metabolic pathways developed by living organisms to ensure their basic metabolic needs and survive in diverse environments.

We developed methods for discovering metabolic pathways from a set of functionally related enzymes (e.g. co-expressed, found in the same operon, regulon or phylogenetic profile group). Those methods rely on graph analysis methods: we build a "generic" metabolic network rasembling all the known reactions, their substrates and products, irrespective of their pertenance to known pathways. This network can then be filtered by selecting organism-specific sub)networks, or weighted according to various criteria (e.g. level of expression of each enzyme measured by transcriptomics methods, exstence of an enzyme in a species or a taxonomic group, ...). We then extract a subgraph that connects at best a set of seed enzymes. We validated this approach by evaluating its capability to recover intermediate reactions of known pathways when only a few seed reactions were provided.

Our current focus is to apply those methods to reconstruct metabolic pathways from bacterial genomes (EU-funded MICROME project), and to analyze metabolic regulation (inferring metabolic pathways from clusters of co-expressed genes).

Perturbation of interactome networks in acute lymphoblastic leukemia

Acute lymphoblastic leukemia (ALL) is the most common cancer in children, and a major cause of mortality among hematopoietic cancers in adults. Despite the significant improvement in the success rate of chemotherapy in recent decades, the rate of recurrence is high and it is urgent to develop alternative therapies that target specific molecules or cellular pathways to improve the chances of cure and reducing toxicity of treatments. The emergence of new technologies has led to many advances in the identification of mutations associated with ALL, but the effects of these mutations on the products of mutated genes and their interactions remain largely unknown. Changes induced by mutations associated with disease should be regarded as a systemic perspective, where networks of interactions are disrupted by genetic alterations. We are studying how molecular interaction networks are perturbed in ALL through high-throughput yeast two-hybrid, statistical network analysis and expression data. We have built a relational database (http://www.bigre.ulb.ac.be/Users/alln/ALLmutations/index.html) to collect and organize information on genes involved in ALL.