[BC]2 Basel Computational Biology Conference 2010 - Regulation and Control in Biological Systems - June 24-25, 2010: Speakers, Titles and Abstracts

8^th [BC]² Basel Computational Biology Conference

Messe Basel

June 24 & 25, 2010

Abstracts

Purely Personal Prejudices Regarding Systems Biology

Marvin Cassman (Marvin Cassman, San Francisco, CA, USA)

Systems biology has moved in just 15 years from an afterthought to a fad. It has, in fact, become the hot label of 21^st century biology, just as “molecular” was the hot label at the end of the 20^th century. The inability or unwillingness to clearly define systems biology, and why it should be considered something different from molecular biology and omics, has resulted in the label being promiscuously applied with some unfortunate consequences. Without a clear understanding of the goals of systems biology it will be difficult to focus on the conceptual and material infrastructure required for further progress.

MicroRNAs as Diagnostics Biomarkers.

Dalia Cohen (Rosetta Genomics, Philadelphia, PA, USA)

MicroRNAs (miRs) are a class of small non- coding RNA molecules that regulate gene expression and therefore have a critical role in many biological processes. MiRs are transcribed and processed from a longer RNA precursor that exhibits a hairpin structure, to a 19-24 nucleotide mature miR. The miR is then incorporated into the RNA-induced silencing complex (RISC), which promotes partial duplex formation between the miR and the 3’ untranslated region (UTR) of targeted mRNAs. This results in translational silencing by either causing mRNA degradation or blocking translation. MiRs have been shown to be involved in normal and pathological conditions and studies indicate significant changes in miR expression in different disease states compared to normal tissue. In cancer, miRs have a major role in tumorigenesis and several miRs have been shown to act as oncogenes and tumor suppressor genes and have unique expression profiles in different malignancies. Importantly, MiRs are stable in clinically relevant samples; they can be detected in fresh and frozen tissues in formalin-fixed, paraffin-embedded (FFPE) tissues and in body fluids. The miRs tissue specific expression and their stability make them ideal biomarkers for disease detection and prognosis.

At Rosetta Genomics we developed proprietary technologies for the detection and quantitation of miRs and by using these platforms, we developed three diagnostic tests. The mirViewTM squamous test enables the differentiation of squamous from non-Squamous non-small cell lung cancer (NSCLC). Current methods used to diagnose NSCLC are limited and offer low accuracy. Furthermore, with the available targeted therapies the precise sub-classification of NSCLC is essential. Certain angiogenic inhibitors that are more effective in the treatment of adenocarcinomas are less effective in treating squamous and are also associated with higher rates of life-threatening pulmonary hemorrhage in this sub-class. The mirViewTM mets test allows the identification of the primary origin of metastases. Current methods for diagnosing origin of metastases are time-consuming, costly and inefficient. The test offers an accurate and efficient diagnostic tool allowing the oncologist to treat the patients with the best treatment modalities.The third test that we developed is mirViewTM meso, which allows the differentiation of mesothelioma from carcinomas in the lung and pleura. This is the first single conclusive test for this important differentiation which provides an important addition to the currently available tools used by pathologists to diagnose this cancer.

In summary, by the unique nature of microRNAs as biomarkers, they are addressing one the most unmet medical needs of today which is personalized medicine and assist the physician in selecting the best treatment for his patient.

Integrated Analysis and Causal Modeling of Complex Biological Networks Defines Biological Mechanisms

Keith O. Elliston (Co-founder and Chairman of the SAB, Genstruct Inc., President, Seneca Creek Research Inc.)

Genetics and Genomics provide very large and coherent datasets that describe complex biological phenomena in terms of genetics and gene expression. These genomic functions result in complex protein and metabolic functions, that when taken in aggregate comprise a biological system. Systems biology attempts to integrate all the key molecular components of a biological system into a single, coherent model that can be used to define biological function, and ultimately to make specific predictions. While defining a ‘steady state’ model for a biological system that can then be perturbed and studied is a long-term goal of systems biology, such technology is still very nascent. In this presentation, we will address the development of large-scale, causal models that define the network that changes in a biological system upon perturbation. These models rely upon the measurement of molecular changes between the base case (control) to the perturbed case (test), and map these changes into a molecular network of cause and effect relations that define the response to the pertubation. These changes include changes in gene expression, genetic mutations, protein concentration and modification, and metabolite concentrations. The causal network model that results from this analysis can be quantified, defining the extent of perturbation of the various network elements, and facilitating a quantitative assessment of the effects of various pertubations on specific elements of the system. This causal network modeling methodology has been used extensively to study the mechanistic affects of drugs and toxins in a variety of biological systems, and forms the basis for the mechanistic modeling of complex biological phenomena in response to pertubations.

Genetic Control of Organ Size in Drosophila

Baumgartner R., Pörnbacher I., Schittenhelm R., Glatter T., Gstaiger M., Stocker H., and Ernst Hafen (Institute of Molecular Systems Biology, ETH Zürich, Switzerland)

Little is known about the mechanisms controlling the size of organs and entire organisms. To identify genes involved in organ size control we conducted a large genetic screen using chemical mutagenesis to achieve an approximately 10-fold saturation of the Drosophila genome. In this way, we have identified some 60 genes affecting head size. The encoded products are involved in a variety of cellular processes including protein synthesis, endocytosis, cell adhesion and nutrient sensing. Two signaling pathways are prominent targets. The insulin pathways plays a central role in controlling cell and organ size in response to nutrient availability. The HIPPO pathway is required for arresting growth when the final organ size is reached. We will describe the use of a combination of genetics and proteomics approaches to map the regulatory network involved in organ size control.

Retrofitting Complex Cellular Systems for the production of biofuels and biochemicals: in need of an ORACLE

Vassily Hatzimanikatis, Keng Cher Soh, Ho Ki Fung, Paolo Angelino, Ljubisa Miskovic (Laboratory of Computational Systems Biotechnology Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland)

The engineering of cells for the production of fuels and chemicals involves simultaneous optimization of multiple objectives, such as specific productivity, extended substrate range and improved tolerance – all under a great degree of uncertainty. The achievement of these objectives under physiological and process constraints will be impossible without the use of mathematical modeling. However, the limited information and the uncertainty in the available information require new methods for modeling and simulation that will characterize the uncertainty and will quantify, in a statistical sense, the expectations of success of alternative metabolic engineering strategies. We discuss these considerations around the development of a framework for the Optimization and Risk Analysis of Complex Living Entities (ORACLE).

Quantitative and Predictive Models of Cellular Signaling

Dagmar Iber (Dept. Biosystems Science and Engineering, ETH Zurich, & SIB, Basel,Switzerland)

Cellular signaling is complex and involves many feedback loops whose role and relevance are difficult to understand by experiments alone. We use a range of modelling techniques to explore signaling network dynamics based on detailed experimental data. In my lecture I will discuss prokaryotic signaling networks that regulate the bacterial stress response as well as the eukaryotic signaling networks that regulate important developmental processes and cell migration.

Integration of Pathway Maps with Empirical ‘Cue-Signal-Response’ Data for Insights Concerning Epithelial Cell Dysregulation

Douglas A. Lauffenburger (Department of Biological Engineering, MIT, Cambridge, MA, USA)

Cellular phenotypic functional behaviors are essentially context-sensitive, so that the dynamic operation of canonical signaling pathways differs from cell type to cell type, between normal and diseased states, and among varied environmental conditions. To date, data-driven pathway reconstructions have largely emphasized production of “averaged” models that hold some general illustrative validity but do not typically provide predictive understanding of cell behavior under different treatments. Effective use of network information in the study of disease and therapy requires context-related models for cells and tissues discerning between normal and dysregulated states. Among featured application areas, we have begun to employ a quantitative, integrative multi-variate proteomic analysis of how signaling networks are dysregulated in epithelial cell pathophysiology, with examples including inflammatory disease and cancer.

Understanding biological systems through metabolic profiling

Ivan Montoliu, François-Pierre Martin, Serge Rezzi (Nestlé Research Center Lausanne, Switzerland)

Metabonomics offers an interesting way to describe and provide new insights into the modulation of regulatory processes in complex biological systems. Overall, it gives a snapshot of the combined effects of genetics and environmental factors that are reflected through complex variations of the global metabolic profile.

The introduction of Nutrimetabonomics has provided a way to gain better understanding on the regulation of metabolic pathways by nutrition. Metabolic profiling is achieved using state-of-the-art analytical methods such as ¹H Nuclear Magnetic Resonance (NMR) and Mass Spectrometry (MS) methods applied to biofluids and tissues. However, changes experienced by the complex biological systems under nutritional interventions are often subtle. Therefore it is imperative to use appropriate data analysis to extract relevant biomarkers from the high density metabolic profiles.

The recent incorporation of algorithms both from chemometrics and machine learning has revealed new opportunities and perspectives for metabolic profile data evaluation. Multi-compartmental metabolic profiling combined with these new approaches is now providing good opportunities to explore the multi-dimensional metabolic connections between various biological matrices. This approach delivers an overview of functional relationships across matrices and enables the characterization of compartment-specific metabolite signatures. Moreover, the approach can be extended to simultaneous metabolic exploration of biological regions of the same system to highlight topographical differences. Highlighting novel metabolic features often comes along with new questions about the underlying biological mechanisms. Moreover, finding a link between highlighted metabolites is often not an easy task. In this scenario, the incorporation of a post-processing modeling step is becoming an essential part of generating valid hypotheses about the origin of metabolic changes.

Rhythmic protein-DNA interactomes and precision in circadian transcription regulatory networks

Felix Naef (EPF Lausanne & SIB)

Circadian oscillators control our daily rhythms in physiology and behavior. These clocks use transcriptional feedback loops that use several key transcription regulators among which the master clock regulator BMAL1. To start dissecting the hierarchical network of transcription regulators behind circadian physiology in mouse liver, we have undertaken a time series ChIP-seq analysis of BMAL1. These experiments also allow us to refine mechanisms for BMAL1/CLOCK dependent transcription, and reveal the dynamic nature of mammalian circadian protein-DNA interactomes on a genome-wide scale I will also present our recent data on the precision of the circadian oscillator that use a novel short lived luciferase reporter. This approach allows us to measure the fine kinetics of transcription from random and circadian promoters in mammalian cells. Analysis in terms of a stochastic model of gene expression gives novel insights into the properties of transcriptional bursting at mammalian promoters.

Motif Activity Response Analysis: Inferring genome-wide transcription regulation in mammals

Erik van Nimwegen (Biozentrum University of Basel & SIB Swiss Institute of Bioinformatics, Basel, Switzerland)

I will discuss an integrated computational approach, called motif activity response analysis (MARA), for reconstructing transcription regulatory networks in mammals from genome-wide expression data. Based on deep sequencing data of transcription start sites we obtained a comprehensive 'promoteromes' in human and mouse, and using probabilistic comparative genomic methods we predict binding sites for over 200 regulatory motifs in proximal promoters genome-wide. Motif Activity Response Analysis (MARA) models genome-wide gene expression profiles in terms of these predicted regulatory sites and I will describe how MARA identifies, for a given system of study, the key regulators driving expression changes, their activity profiles across the samples, and the sets of target promoters of each regulator. Time permitting I will talk about how MARA can be extended to incorporate epigenetic changes to chromatin structure.

Algorithmic Systems Biology

Corrado Priami (The Microsoft Research - University of Trento Centre for Computational and Systems Biology, Italy)

The convergence between computer science and biology occurred in successive waves, involving deeper and deeper concepts of computing. The current situation makes computer science a suitable candidate for becoming a philosophical foundation for systems biology with the same importance as mathematics, chemistry and physics. However, this significant opportunity is not a free lunch. New developments and a strong integration of different fields of computing are needed to face the challenges of systems biology. One of these developments is that of a complex and expanding applicative domain that can open entirely new avenues of research in computing and eventually help it become a natural, quantitative science. More info in "Algorithmic Systems Biology", Communications of the ACM, 52(5):80-88, May 2009.

"Algorithmic Systems Biology", Communications of the ACM, 52(5):80-88, May 2009. [Weblink]

Transcriptional control of metabolic fluxes and computational identification of the governing principles

Roelco Kleijn, Sarah-Maria Fendt, Robert Schutz, Matthias Heinemann, Nicola Zamboni and Uwe Sauer (Institute of Molecular Systems Biology, ETH Zurich, Switzerland)

Great strides have been made in our ability to monitor transcriptional and proteome responses, but prediction of biological function and activity from such data has remained challenging (1). Here we elucidate which transcription factors actually control metabolic function in a given environment on the basis of large-scale flux analyses (2) upon genetic and environmental perturbations in E. coli and B. subtilis. Systematic flux analysis of 120 transcription factor mutants in the yeast S. cerevisiae then reveal networks of active transcriptional regulation under 5 conditions. Using metabolomics and targeted proteomics for in depth analysis of key mutants then identify the precise molecular targets of these active regulation mechanisms. While many proteins are differentially expressed in deletions mutants and under different conditions, only relatively few of these expression changes actually cause flux alterations, as revealed by computational analyses of flux, proteome and metabolome data.

On the basis of these large in vivo flux data, we then ask whether there are generally valid principles that describe the distribution of flux under different conditions and how such metabolic networks respond to perturbations? Using the computational framework of flux balance analysis, we test two fundamentally different families of hypotheses: are cells optimized during evolution towards one or more objectives (3) or are their responses optimized towards minimal readjustments?

Heinemann M & Sauer U. Curr. Opin. Microbiol. 2010 Mar.
Fischer, E. & Sauer, U. Nat. Genet. 37, 636-640 (2005).
R Schütz, L Küpfer & U Sauer. Mol. Sys. Biol. 3:119 (2007).

Uncovering the Human cell lineage tree in health and disease: The next grand scientific challenge

Ehud Shapiro (Weizmann Institute of Science, Rehovot, Israel)

The cell lineage tree of a person captures the history of the person’s cells since conception. In computer science terms it is a rooted, labeled binary tree, where the root represents the primary fertilized egg, leaves represent extant cells, internal nodes represent past cell divisions, and vertex labels record cell types. It has approximately 100 trillion leaves and 100 trillion branches (≈100,000 bigger than the Human genome); it is unknown. We should strive to know it, as many central questions in biology and medicine are actually specific questions about the Human cell lineage tree, in health and disease: Which cancer cells initiate relapse after chemotherapy? Which cancer cells can metastasize? Do insulin-producing beta cells renew in healthy adults? Do eggs renew in adult females? Which cells renew in healthy and in unhealthy adult brain? Knowing the Human cell lineage tree would answer all these questions and more.

Fortunately, our cell lineage tree is implicitly encoded in our cells’ genomes via mutations that accumulate when body cells divide. Theoretically, it could be reconstructed with high precision by sequencing every cell in our body, at a prohibitive cost. Practically, analyzing only highly-mutable fragments of the genome is sufficient for cell lineage reconstruction. Our lab has developed a proof-of-concept multi-disciplinary method and system for cell lineage analysis from somatic mutations. The talk will describe the system and results obtained with it so far, and a proposal for a FET Flagship project for uncovering the Human cell lineage tree in health and disease.

Systems analysis of cellular regulation under uncertainty

Jörg Stelling (Dept. Biosystems Science and Engineering, ETH Zurich, & SIB, Basel, Switzerland)

For complex cellular networks, limited mechanistic knowledge, conflicting hypotheses, and relatively scarce experimental data hamper the development of mathematical models as systems analysis tools. The talk focuses on two approaches for dealing with this combination of complexity and uncertainty. They combine theory development and applications to specific biological examples. Firstly, reaction network stoichiometries are relatively well-characterized and therefore suitable starting points for the analysis. Beyond the traditional applications to metabolic networks, theory extensions connect structural network analysis to systems dynamics, for instance, to identify key mechanisms in cellular decision processes. Secondly, and more mechanistically, we address uncertain biological mechanisms by casting hypotheses into a library of dynamic mathematical models, evaluating these against experimental observations, and designing pivotal experiments to discriminate between alternatives. This approach proved fruitful in elucidating key signalling mechanism for the Target of Rapamycin (TOR) pathway in yeast as well as for more generic analyses of the relations between promoter complexity and performance. Overall, these studies stress the importance of network structures, which seems to outweigh the fine tuning of parameters. Structure-oriented analysis of biological systems, thus, provides challenging theory problems as well as broad perspectives for uncovering the organization and functionality of cellular networks.

Uncovering miRNA-dependent post-transcriptional regulatory networks

Mihaela Zavolan (Biozentrum University of Basel & SIB Swiss Institute of Bioinformatics, Basel, Switzerland)

MiRNAs are 21-23 nucleotides-long RNAs that are involved in a broad range of biological processes, from cell division and metabolism, to development and immune responses. They guide the RNA-induced silencing complex (RISC) to mRNA targets which, upon RISC binding undergo translational inhibition, decapping, deadenylation and degradation. How the specificity of miRNA-target interaction emerges is not entirely known, and it appears to require relatively little sequence complementarity. In this presentation I will present our efforts to identify and characterize miRNA target sites, to uncover the determinants of miRNA-target interactions and to evaluate the role of miRNAs in the regulation of cellular processes.

Modeling spatio-temporal-response expression in multiple species

Philip Zimmermann (Genevestigator / ETH Zürich, Switzerland)

High-throughput molecular profiling of biological samples has become a mainstream approach in biomedical research. Nevertheless, the massive amount of data arising from this research often remains under-analyzed. Frequently, single experiments are analyzed in detail but in isolation from the existing body of data. The next logical step is to build systems that allow to integrate and analyze thousands of experiments simultaneously and thereby to identify regulatory relationships between genes and a broad variety of conditions. A prerequisite to this type of analysis is the systematic sample annotation, quality control, and normalization of the data. Over several years, the Genevestigator team has curated manually more than 40,000 public Affymetrix arrays, creating a unique and high quality database of expression data. The use of ontologies to annotate tissue type, development, and perturbation (e.g. disease, chemical treatment, tumor) makes it possible to model expression in the dimensions of space, time and response. Our team is strongly interested in creating methods and tools to understand the spatio-temporal-response regulation of genes and networks, and to identify biomarker genes with precisely defined spatio-temporal-response profiles. Concepts and results from this research will be presented.


Home Registration Info Program Sponsors
Add to my agenda: [ vcs ] [ ics ]		8^th [BC]² Basel Computational Biology Conference Messe Basel June 24 & 25, 2010 Abstracts Purely Personal Prejudices Regarding Systems Biology Marvin Cassman (Marvin Cassman, San Francisco, CA, USA) Systems biology has moved in just 15 years from an afterthought to a fad. It has, in fact, become the hot label of 21^st century biology, just as “molecular” was the hot label at the end of the 20^th century. The inability or unwillingness to clearly define systems biology, and why it should be considered something different from molecular biology and omics, has resulted in the label being promiscuously applied with some unfortunate consequences. Without a clear understanding of the goals of systems biology it will be difficult to focus on the conceptual and material infrastructure required for further progress. MicroRNAs as Diagnostics Biomarkers. Dalia Cohen (Rosetta Genomics, Philadelphia, PA, USA) MicroRNAs (miRs) are a class of small non- coding RNA molecules that regulate gene expression and therefore have a critical role in many biological processes. MiRs are transcribed and processed from a longer RNA precursor that exhibits a hairpin structure, to a 19-24 nucleotide mature miR. The miR is then incorporated into the RNA-induced silencing complex (RISC), which promotes partial duplex formation between the miR and the 3’ untranslated region (UTR) of targeted mRNAs. This results in translational silencing by either causing mRNA degradation or blocking translation. MiRs have been shown to be involved in normal and pathological conditions and studies indicate significant changes in miR expression in different disease states compared to normal tissue. In cancer, miRs have a major role in tumorigenesis and several miRs have been shown to act as oncogenes and tumor suppressor genes and have unique expression profiles in different malignancies. Importantly, MiRs are stable in clinically relevant samples; they can be detected in fresh and frozen tissues in formalin-fixed, paraffin-embedded (FFPE) tissues and in body fluids. The miRs tissue specific expression and their stability make them ideal biomarkers for disease detection and prognosis. At Rosetta Genomics we developed proprietary technologies for the detection and quantitation of miRs and by using these platforms, we developed three diagnostic tests. The mirViewTM squamous test enables the differentiation of squamous from non-Squamous non-small cell lung cancer (NSCLC). Current methods used to diagnose NSCLC are limited and offer low accuracy. Furthermore, with the available targeted therapies the precise sub-classification of NSCLC is essential. Certain angiogenic inhibitors that are more effective in the treatment of adenocarcinomas are less effective in treating squamous and are also associated with higher rates of life-threatening pulmonary hemorrhage in this sub-class. The mirViewTM mets test allows the identification of the primary origin of metastases. Current methods for diagnosing origin of metastases are time-consuming, costly and inefficient. The test offers an accurate and efficient diagnostic tool allowing the oncologist to treat the patients with the best treatment modalities.The third test that we developed is mirViewTM meso, which allows the differentiation of mesothelioma from carcinomas in the lung and pleura. This is the first single conclusive test for this important differentiation which provides an important addition to the currently available tools used by pathologists to diagnose this cancer. In summary, by the unique nature of microRNAs as biomarkers, they are addressing one the most unmet medical needs of today which is personalized medicine and assist the physician in selecting the best treatment for his patient. Integrated Analysis and Causal Modeling of Complex Biological Networks Defines Biological Mechanisms Keith O. Elliston (Co-founder and Chairman of the SAB, Genstruct Inc., President, Seneca Creek Research Inc.) Genetics and Genomics provide very large and coherent datasets that describe complex biological phenomena in terms of genetics and gene expression. These genomic functions result in complex protein and metabolic functions, that when taken in aggregate comprise a biological system. Systems biology attempts to integrate all the key molecular components of a biological system into a single, coherent model that can be used to define biological function, and ultimately to make specific predictions. While defining a ‘steady state’ model for a biological system that can then be perturbed and studied is a long-term goal of systems biology, such technology is still very nascent. In this presentation, we will address the development of large-scale, causal models that define the network that changes in a biological system upon perturbation. These models rely upon the measurement of molecular changes between the base case (control) to the perturbed case (test), and map these changes into a molecular network of cause and effect relations that define the response to the pertubation. These changes include changes in gene expression, genetic mutations, protein concentration and modification, and metabolite concentrations. The causal network model that results from this analysis can be quantified, defining the extent of perturbation of the various network elements, and facilitating a quantitative assessment of the effects of various pertubations on specific elements of the system. This causal network modeling methodology has been used extensively to study the mechanistic affects of drugs and toxins in a variety of biological systems, and forms the basis for the mechanistic modeling of complex biological phenomena in response to pertubations. Genetic Control of Organ Size in Drosophila Baumgartner R., Pörnbacher I., Schittenhelm R., Glatter T., Gstaiger M., Stocker H., and Ernst Hafen (Institute of Molecular Systems Biology, ETH Zürich, Switzerland) Little is known about the mechanisms controlling the size of organs and entire organisms. To identify genes involved in organ size control we conducted a large genetic screen using chemical mutagenesis to achieve an approximately 10-fold saturation of the Drosophila genome. In this way, we have identified some 60 genes affecting head size. The encoded products are involved in a variety of cellular processes including protein synthesis, endocytosis, cell adhesion and nutrient sensing. Two signaling pathways are prominent targets. The insulin pathways plays a central role in controlling cell and organ size in response to nutrient availability. The HIPPO pathway is required for arresting growth when the final organ size is reached. We will describe the use of a combination of genetics and proteomics approaches to map the regulatory network involved in organ size control. Retrofitting Complex Cellular Systems for the production of biofuels and biochemicals: in need of an ORACLE Vassily Hatzimanikatis, Keng Cher Soh, Ho Ki Fung, Paolo Angelino, Ljubisa Miskovic (Laboratory of Computational Systems Biotechnology Ecole Polytechnique Federale de Lausanne (EPFL), Switzerland) The engineering of cells for the production of fuels and chemicals involves simultaneous optimization of multiple objectives, such as specific productivity, extended substrate range and improved tolerance – all under a great degree of uncertainty. The achievement of these objectives under physiological and process constraints will be impossible without the use of mathematical modeling. However, the limited information and the uncertainty in the available information require new methods for modeling and simulation that will characterize the uncertainty and will quantify, in a statistical sense, the expectations of success of alternative metabolic engineering strategies. We discuss these considerations around the development of a framework for the Optimization and Risk Analysis of Complex Living Entities (ORACLE). Quantitative and Predictive Models of Cellular Signaling Dagmar Iber (Dept. Biosystems Science and Engineering, ETH Zurich, & SIB, Basel,Switzerland) Cellular signaling is complex and involves many feedback loops whose role and relevance are difficult to understand by experiments alone. We use a range of modelling techniques to explore signaling network dynamics based on detailed experimental data. In my lecture I will discuss prokaryotic signaling networks that regulate the bacterial stress response as well as the eukaryotic signaling networks that regulate important developmental processes and cell migration. Integration of Pathway Maps with Empirical ‘Cue-Signal-Response’ Data for Insights Concerning Epithelial Cell Dysregulation Douglas A. Lauffenburger (Department of Biological Engineering, MIT, Cambridge, MA, USA) Cellular phenotypic functional behaviors are essentially context-sensitive, so that the dynamic operation of canonical signaling pathways differs from cell type to cell type, between normal and diseased states, and among varied environmental conditions. To date, data-driven pathway reconstructions have largely emphasized production of “averaged” models that hold some general illustrative validity but do not typically provide predictive understanding of cell behavior under different treatments. Effective use of network information in the study of disease and therapy requires context-related models for cells and tissues discerning between normal and dysregulated states. Among featured application areas, we have begun to employ a quantitative, integrative multi-variate proteomic analysis of how signaling networks are dysregulated in epithelial cell pathophysiology, with examples including inflammatory disease and cancer. Understanding biological systems through metabolic profiling Ivan Montoliu, François-Pierre Martin, Serge Rezzi (Nestlé Research Center Lausanne, Switzerland) Metabonomics offers an interesting way to describe and provide new insights into the modulation of regulatory processes in complex biological systems. Overall, it gives a snapshot of the combined effects of genetics and environmental factors that are reflected through complex variations of the global metabolic profile. The introduction of Nutrimetabonomics has provided a way to gain better understanding on the regulation of metabolic pathways by nutrition. Metabolic profiling is achieved using state-of-the-art analytical methods such as ¹H Nuclear Magnetic Resonance (NMR) and Mass Spectrometry (MS) methods applied to biofluids and tissues. However, changes experienced by the complex biological systems under nutritional interventions are often subtle. Therefore it is imperative to use appropriate data analysis to extract relevant biomarkers from the high density metabolic profiles. The recent incorporation of algorithms both from chemometrics and machine learning has revealed new opportunities and perspectives for metabolic profile data evaluation. Multi-compartmental metabolic profiling combined with these new approaches is now providing good opportunities to explore the multi-dimensional metabolic connections between various biological matrices. This approach delivers an overview of functional relationships across matrices and enables the characterization of compartment-specific metabolite signatures. Moreover, the approach can be extended to simultaneous metabolic exploration of biological regions of the same system to highlight topographical differences. Highlighting novel metabolic features often comes along with new questions about the underlying biological mechanisms. Moreover, finding a link between highlighted metabolites is often not an easy task. In this scenario, the incorporation of a post-processing modeling step is becoming an essential part of generating valid hypotheses about the origin of metabolic changes. Rhythmic protein-DNA interactomes and precision in circadian transcription regulatory networks Felix Naef (EPF Lausanne & SIB) Circadian oscillators control our daily rhythms in physiology and behavior. These clocks use transcriptional feedback loops that use several key transcription regulators among which the master clock regulator BMAL1. To start dissecting the hierarchical network of transcription regulators behind circadian physiology in mouse liver, we have undertaken a time series ChIP-seq analysis of BMAL1. These experiments also allow us to refine mechanisms for BMAL1/CLOCK dependent transcription, and reveal the dynamic nature of mammalian circadian protein-DNA interactomes on a genome-wide scale I will also present our recent data on the precision of the circadian oscillator that use a novel short lived luciferase reporter. This approach allows us to measure the fine kinetics of transcription from random and circadian promoters in mammalian cells. Analysis in terms of a stochastic model of gene expression gives novel insights into the properties of transcriptional bursting at mammalian promoters. Motif Activity Response Analysis: Inferring genome-wide transcription regulation in mammals Erik van Nimwegen (Biozentrum University of Basel & SIB Swiss Institute of Bioinformatics, Basel, Switzerland) I will discuss an integrated computational approach, called motif activity response analysis (MARA), for reconstructing transcription regulatory networks in mammals from genome-wide expression data. Based on deep sequencing data of transcription start sites we obtained a comprehensive 'promoteromes' in human and mouse, and using probabilistic comparative genomic methods we predict binding sites for over 200 regulatory motifs in proximal promoters genome-wide. Motif Activity Response Analysis (MARA) models genome-wide gene expression profiles in terms of these predicted regulatory sites and I will describe how MARA identifies, for a given system of study, the key regulators driving expression changes, their activity profiles across the samples, and the sets of target promoters of each regulator. Time permitting I will talk about how MARA can be extended to incorporate epigenetic changes to chromatin structure. Algorithmic Systems Biology Corrado Priami (The Microsoft Research - University of Trento Centre for Computational and Systems Biology, Italy) The convergence between computer science and biology occurred in successive waves, involving deeper and deeper concepts of computing. The current situation makes computer science a suitable candidate for becoming a philosophical foundation for systems biology with the same importance as mathematics, chemistry and physics. However, this significant opportunity is not a free lunch. New developments and a strong integration of different fields of computing are needed to face the challenges of systems biology. One of these developments is that of a complex and expanding applicative domain that can open entirely new avenues of research in computing and eventually help it become a natural, quantitative science. More info in "Algorithmic Systems Biology", Communications of the ACM, 52(5):80-88, May 2009. "Algorithmic Systems Biology", Communications of the ACM, 52(5):80-88, May 2009. [Weblink] Transcriptional control of metabolic fluxes and computational identification of the governing principles Roelco Kleijn, Sarah-Maria Fendt, Robert Schutz, Matthias Heinemann, Nicola Zamboni and Uwe Sauer (Institute of Molecular Systems Biology, ETH Zurich, Switzerland) Great strides have been made in our ability to monitor transcriptional and proteome responses, but prediction of biological function and activity from such data has remained challenging (1). Here we elucidate which transcription factors actually control metabolic function in a given environment on the basis of large-scale flux analyses (2) upon genetic and environmental perturbations in E. coli and B. subtilis. Systematic flux analysis of 120 transcription factor mutants in the yeast S. cerevisiae then reveal networks of active transcriptional regulation under 5 conditions. Using metabolomics and targeted proteomics for in depth analysis of key mutants then identify the precise molecular targets of these active regulation mechanisms. While many proteins are differentially expressed in deletions mutants and under different conditions, only relatively few of these expression changes actually cause flux alterations, as revealed by computational analyses of flux, proteome and metabolome data. On the basis of these large in vivo flux data, we then ask whether there are generally valid principles that describe the distribution of flux under different conditions and how such metabolic networks respond to perturbations? Using the computational framework of flux balance analysis, we test two fundamentally different families of hypotheses: are cells optimized during evolution towards one or more objectives (3) or are their responses optimized towards minimal readjustments? Heinemann M & Sauer U. Curr. Opin. Microbiol. 2010 Mar. Fischer, E. & Sauer, U. Nat. Genet. 37, 636-640 (2005). R Schütz, L Küpfer & U Sauer. Mol. Sys. Biol. 3:119 (2007). Uncovering the Human cell lineage tree in health and disease: The next grand scientific challenge Ehud Shapiro (Weizmann Institute of Science, Rehovot, Israel) The cell lineage tree of a person captures the history of the person’s cells since conception. In computer science terms it is a rooted, labeled binary tree, where the root represents the primary fertilized egg, leaves represent extant cells, internal nodes represent past cell divisions, and vertex labels record cell types. It has approximately 100 trillion leaves and 100 trillion branches (≈100,000 bigger than the Human genome); it is unknown. We should strive to know it, as many central questions in biology and medicine are actually specific questions about the Human cell lineage tree, in health and disease: Which cancer cells initiate relapse after chemotherapy? Which cancer cells can metastasize? Do insulin-producing beta cells renew in healthy adults? Do eggs renew in adult females? Which cells renew in healthy and in unhealthy adult brain? Knowing the Human cell lineage tree would answer all these questions and more. Fortunately, our cell lineage tree is implicitly encoded in our cells’ genomes via mutations that accumulate when body cells divide. Theoretically, it could be reconstructed with high precision by sequencing every cell in our body, at a prohibitive cost. Practically, analyzing only highly-mutable fragments of the genome is sufficient for cell lineage reconstruction. Our lab has developed a proof-of-concept multi-disciplinary method and system for cell lineage analysis from somatic mutations. The talk will describe the system and results obtained with it so far, and a proposal for a FET Flagship project for uncovering the Human cell lineage tree in health and disease. Systems analysis of cellular regulation under uncertainty Jörg Stelling (Dept. Biosystems Science and Engineering, ETH Zurich, & SIB, Basel, Switzerland) For complex cellular networks, limited mechanistic knowledge, conflicting hypotheses, and relatively scarce experimental data hamper the development of mathematical models as systems analysis tools. The talk focuses on two approaches for dealing with this combination of complexity and uncertainty. They combine theory development and applications to specific biological examples. Firstly, reaction network stoichiometries are relatively well-characterized and therefore suitable starting points for the analysis. Beyond the traditional applications to metabolic networks, theory extensions connect structural network analysis to systems dynamics, for instance, to identify key mechanisms in cellular decision processes. Secondly, and more mechanistically, we address uncertain biological mechanisms by casting hypotheses into a library of dynamic mathematical models, evaluating these against experimental observations, and designing pivotal experiments to discriminate between alternatives. This approach proved fruitful in elucidating key signalling mechanism for the Target of Rapamycin (TOR) pathway in yeast as well as for more generic analyses of the relations between promoter complexity and performance. Overall, these studies stress the importance of network structures, which seems to outweigh the fine tuning of parameters. Structure-oriented analysis of biological systems, thus, provides challenging theory problems as well as broad perspectives for uncovering the organization and functionality of cellular networks. Uncovering miRNA-dependent post-transcriptional regulatory networks Mihaela Zavolan (Biozentrum University of Basel & SIB Swiss Institute of Bioinformatics, Basel, Switzerland) MiRNAs are 21-23 nucleotides-long RNAs that are involved in a broad range of biological processes, from cell division and metabolism, to development and immune responses. They guide the RNA-induced silencing complex (RISC) to mRNA targets which, upon RISC binding undergo translational inhibition, decapping, deadenylation and degradation. How the specificity of miRNA-target interaction emerges is not entirely known, and it appears to require relatively little sequence complementarity. In this presentation I will present our efforts to identify and characterize miRNA target sites, to uncover the determinants of miRNA-target interactions and to evaluate the role of miRNAs in the regulation of cellular processes. Modeling spatio-temporal-response expression in multiple species Philip Zimmermann (Genevestigator / ETH Zürich, Switzerland) High-throughput molecular profiling of biological samples has become a mainstream approach in biomedical research. Nevertheless, the massive amount of data arising from this research often remains under-analyzed. Frequently, single experiments are analyzed in detail but in isolation from the existing body of data. The next logical step is to build systems that allow to integrate and analyze thousands of experiments simultaneously and thereby to identify regulatory relationships between genes and a broad variety of conditions. A prerequisite to this type of analysis is the systematic sample annotation, quality control, and normalization of the data. Over several years, the Genevestigator team has curated manually more than 40,000 public Affymetrix arrays, creating a unique and high quality database of expression data. The use of ontologies to annotate tissue type, development, and perturbation (e.g. disease, chemical treatment, tumor) makes it possible to model expression in the dimensions of space, time and response. Our team is strongly interested in creating methods and tools to understand the spatio-temporal-response regulation of genes and networks, and to identify biomarker genes with precisely defined spatio-temporal-response profiles. Concepts and results from this research will be presented.

The [BC]² Basel Computational Biology Conference is a symposium of the SIB Swiss Institute of Bioinformatics organized by:		[Contact] [Center] [About]