Invited speakers
- Céline Brochier (Université Lyon 1, France)
The growing tree of Archaea: changing perspectives on the diversity and evolution of the third domain of life
[abstract] - Franca Fraternali (Randall Division of Cellular and Molecular Biophysics, King’s college, London, United Kingdom)
Unraveling the Good and the Bad in Protein Networks: Functional versus Dysfunctional Interactions.
[abstract] - John Huelsenbeck (Department of Integrative Biology, UC Berkeley, USA)
Bayesian inference in phylogeny for genome-scale data.
[abstract] - Tobias Marschall (Algorithms for Computational Genomics, Max-Planck-Institut für Informatik, Saabrücken, Germany)
A Guided Tour to Computational Haplotyping.
[abstract] - Julio Saez-Rodriguez (RWTH, Aachen University, Germany and EMBL-EBI)
Network Models to Understand and Combat Cancer: From Clinical Genomics to Biochemical Modelling.
[abstract] - Patrick Wincker (CEA Genoscope, Evry, France)
Holistic metagenomics in marine plankton communities
The growing tree of Archaea: changing perspectives on the diversity and evolution of the third domain of life.
Archaea occupy a key position in the Tree of Life, and represent a major fraction of the microbial diversity. Abundant in soils, ocean sediments and the water column, they are key players in processes mediating global carbon and nutrient fluxes, as well as important components of the animal microbiome and human body. The development of culture-independent sequencing techniques has revealed a myriad of so far inaccessible microbial lineages and filled up the archaeal tree with entirely new branches.
The unprecedented access to genomic data from a large number of archaeal lineages provides the raw material for dissecting the origin of this domain, the evolutionary trajectories that have shaped its current diversity, and its relationships with Bacteria and Eukaryotes. This rainfall of data combined to cutting-edge methods allowing to disentangle the multiple signals contained in molecular sequences has shed new light on the evolutionary history of Archaea.
Here I will review the major advances in the field as well as important open issues and future challenges.
Unraveling the Good and the Bad in Protein Networks: Functional versus Dysfunctional Interactions.
In the last years, protein interactome comparisons have highlighted conserved modules that might represent common functional cores of ancestral origin. However, recent analyses of protein-protein interaction networks (PPINs) have led to a debate about the influence of the experimental method on the quality and biological relevance of these interaction data. It is crucial to know to what extent discrepancies between the networks of different species reflect sampling biases of the respective experimental methods, as opposed to topological features due to biological functionality. This requires new, precise and practical mathematical tools to quantify and compare the topological structures of networks at high resolution. To this end, we have studied the relationship between structured random graph ensembles and real biological signaling networks, focusing on the number of short loops in networks, which represent complexes in PPINs. By combination of a method for graph dynamics and an algorithm for loop counting, we estimated the relative importance of loops in biological networks compared to random graphs. We found that loops are a predominant feature of PPINs, suggesting that enrichment of their occurrence has a key functional role.
Nevertheless, one must keep in mind that not all the interactions between proteins result in a functional role that benefits the cell. One example is protein aggregation, resulting in neurotoxic assemblies that lead ultimately to cell death. We investigate in detail the case of interactions between fragments of the Prion protein (PrP) constituted by only the helices H2 and H3 of the entire protein. We have investigated the molecular mechanisms of the self-assembly process in solution by Molecular Dynamics. Our simulations show that this process occurs by assembly of small modules of four monomers that precede the creation of a “base” of six to eight H2H3 monomers; starting from this “base”, other H2H3 units attach to it in various configurations, assembling short filaments.
1) Chung SS, Pandini A, Annibale A, Coolen AC, Thomas NS, Fraternali F. Bridging topological and functional information in protein interaction networks by short loops profiling. Sci Rep. 2015; 5:8540.
2) Chakroun N, Fornili A, Prigent S, Kleinjung J, Dreiss CA, Rezaei H, Fraternali F. Decrypting Prion Protein Conversion into a β-Rich Conformer by Molecular Dynamics. J Chem Theory Comput. 2013; 5:2455-2465.
3) Chakroun N, Prigent S, Dreiss CA, Noinville S, Chapuis C, Fraternali F, Rezaei H. The oligomerization properties of prion protein are restricted to the H2H3 domain. FASEB J. 2010; 9:3222-31.
Bayesian inference in phylogeny for genome-scale data
Bayesian inference has permeated the field of phylogenetics. A major challenge in the field remains how to extend methods to genome-scale data. The temptation is to take short-cuts by applying fast methods that do not take full advantage of the information contained in the data. I describe several methods that may be applicable to genome-scale
data. First, I describe new proposal mechanisms for better inferring large phylogenetic trees. Second, I discuss a class of models that can be used to address questions such as the identification of sites under the influence of natural selection.
A Guided Tour to Computational Haplotyping.
Humans and many other species are diploid. Every individual inherits two versions of each autosomal chromosome, called haplotypes, one from its mother and one from its father. Moving from (sequences of) genotypes to haplotypes is known as phasing or haplotyping. The knowledge of haplotypes is critical for addressing a variety of important questions in fundamental and clinical research. In this talk, I will highlight both algorithmic and experimental aspects of reconstructing haplotypes, with a special emphasis on recent technological advancements and their impact on the computational problems to be solved. I will briefly touch on population-based and pedigree-based phasing method, but will mostly focus on direct experimental methods that allow to reconstruct haplotypes for single individuals. Haplotype reconstruction from sequencing reads is most commonly formalized as the Minimum Error Correction (MEC) problem. Recent advances on fixed-parameter tractable (FPT) algorithm allow us to (quickly) solve practically relevant instances of this NP-hard problem optimally. I will present experimental results from five different platforms (PacBio, Oxford Nanopore, Hi-C, StrandSeq, and 10X Genomics) and highlight how combinations of these technologies allow to accurately reconstruct dense chromosome-length human haplotypes at manageable costs.
Network Models to Understand and Combat Cancer: From Clinical Genomics to Biochemical Modelling.
Large-scale genomic studies are providing unprecedented insights into the molecular basis of cancer, but it remains challenging to leverage this information for the development and application of therapies. We have performed an integrated analysis of the molecular profiles of 11,215 primary tumours and 1,001 cancer cell lines, along with the response of the cell lines to 265 anti-cancer compounds. This analysis finds alterations in tumours that can confer drug sensitivity or resistance, and sheds light on which data types are most informative to prioritize treatment. Integration of this data with various sources of prior knowledge, in particular signaling pathways and transcription factors, points at molecular processes involved in resistance mechanisms, and offer hypotheses for novel combination therapies. Our own analysis as well as the results of a crowdsourcing effort (DREAM challenge) reveals that prediction of drug efficacy is far from accurate, implying important limitations for personalised medicine. I will argue than an important missing aspect is the dynamics of signaling networks, and show how applying logic models, trained with phosphoproteomic measurements upon perturbations, can further improve our understanding of the molecular basis of drug resistance, thereby providing new treatment opportunities not noticeable by static molecular characterization.