Dealing with complexity of biological systems: from data to models

Abstract

Four chapters of the synthesis represent four major areas of my research interests: 1) data analysis in molecular biology, 2) mathematical modeling of biological networks, 3) genome evolution, and 4) cancer systems biology. The first chapter is devoted to my work in developing non-linear methods of dimension reduction (methods of elastic maps and principal trees) which extends the classical method of principal components. Also I present application of matrix factorization techniques to analysis of cancer data. The second chapter is devoted to the complexity of mathematical models in molecular biology. I describe the basic ideas of asymptotology of chemical reaction networks aiming at dissecting and simplifying complex chemical kinetics models. Two applications of this approach are presented: to modeling NFkB and apoptosis pathways, and to modeling mechanisms of miRNA action on protein translation. The third chapter briefly describes my investigations of the genome structure in different organisms (from microbes to human cancer genomes). Unsupervised data analysis approaches are used to investigate the patterns in genomic sequences shaped by genome evolution and influenced by the basic properties of the environment. The fourth chapter summarizes my experience in studying cancer by computational methods (through combining integrative data analysis and mathematical modeling approaches). In particular, I describe the on-going research projects such as mathematical modeling of cell fate decisions and synthetic lethal interactions in DNA repair network. The synthesis is concluded by listing major challenges in computational systems biology, connected to the topics of this text, i.e. dealing with complexity of biological systems.