You are here: TUCS > RESEARCH > Research Units > Computational Biomodeling Laboratory (Combio Lab)
Computational Biomodeling Laboratory (Combio Lab)
The research of the laboratory centers on computational methods for modelling biochemical systems. The general interest of the laboratory is gaining an understanding of the fundamental computational and information-processing principles behind the functioning of bio-systems. We have considerable expertise in building discrete models, based on combinatorics, graph theory, and stochastic processes. We are also experts in evaluating such models against experimental data, discovering their control structure, quantitative model comparison and quantitative model refinement.
The laboratory has hosted in the last 5 years 5 postdoctoral students and has graduated 3 TUCS PhD students, 2 of them receiving their degrees with honors. The scientific volume of the unit has been consistently very good, both in quality and in quantity. The laboratory is actively involved in the editorial boards of several journals and the program committees of the most relevant conference in its field of research.
Research Unit Web Page: http://combio.abo.fi/
Leader of the unit
Ion PetreResearchers
Vladimir Rogojin An Le Thi ThanhDoctoral Students
Bogdan Iancu Sepinoud Azimi Diana-Elena Gratie Charmi PanchalUndergraduate Students
Romina PaturcaProjects
Quantitative strategies for the self-assembly of intermediate filaments
In our research we concentrate on the process of in vitro self-assembly of intermediate filaments from tetrameric vimentin. We investigate different plausible strategies for filament elongation through mathematical modelling, model fitting, model validation and sensitivity analysis. In the assessment of the potential variants the focus is on properties such as scalability, robustness and ability to explain experimental data. This systematic approach enables the formulation of certain hypotheses about how the still little-known process of filament self-assembly is executed. Based on this hypotheses future biological experiments that would verify them are proposed. This project is an example of a hypothesis-driven research in the field of systems biology.
Quantitative model refinement
We focus in this project on computational techniques allowing the quantitative refinement of a model without altering its numerical fit and validation. Our research addresses two main problems in the design of mathematical models in systems biology: (i) the quantitative fit and validation of a large model is a computationally difficult problem; (ii) changing a model (e.g., adding details to it) implies redoing the work on the numerical fit and validation of the model. Our proposed methodology builds on the expertise gained in computer science in (qualitative) program refinement, extending it in a fundamental way to the realm of quantitative biomodels.
Computational modeling of the eukaryotic heat shock response
Cells exposed to elevated temperature or other stress stimuli respond by increased expression of heat shock proteins (HSPs). The heat shock response and the proteins involved have been highly conserved throughout evolution from Escherichia coli to human. In addition to heat, a wide variety of biological (infection, inflammation), physical (radiation, hypoxia) and chemical (alcohols, metals) stressors can induce the response. This is why the heat shock response is also called “stress response” and the heat shock proteins, in consequence, “stress proteins”. We investigate in this project a new molecular model for the heat shock response including the stress-induced response and its self-regulation. We focus on: (i) understanding the main control mechanisms in the architecture of the heat shoch response and (ii) new modeling methods to allow the integration of the phosphorylation-based control mechanisms for the heat shock response, while addressing the problem of the combinatorial explosion of the model size.
Funded by Academy of Finland, 2008-2010.
Computing at nano-scale
We investigate mathematical models for self-assembly, contributing to laying solid foundations for nano-science, that are still missing to a large extent at this time. Based on such foundations, we seek to clarify several central questions: e.g., what can be effectively self-assembled (and thus nano-fabricated), how complex is it to self-assemble a given shape, or what initial structures can self-assemble into a certain shape.
Funded by Academy of Finland, 2005-2010
Computational processes in living cells
The process of gene assembly has the attention of the Biocomputing community for several years already. It is by now clear that the process of gene assembly in ciliates is highly computational: it turns out that ciliates "know" one of the basic data structures of Computer Science - the linked list - and use it in a very elegant pattern matching manner in the process of gene assembly! We are investigating a set of three molecular operations that accomplishes the gene assembly through the "fold and recombine" paradigm. We introduced the mathematical model of pointer reduction systems to formalize the micronuclear gene patterns (through permutations, strings and graphs) and the gene assembly process. Our investigation of these systems resulted in a uniform explanation of all known experimental results concerning gene assembly in ciliates.
Funded by Academy of Finland, 2004-2007, within the research program for Systems Biology and Bioinformatics.
Molecular computing network
Molecular computing is a novel, exciting and a genuinely interdisciplinary research area which lies at the boundary of Computer Science and Molecular Biology. An important advantage offered by computations with bio-molecules is the massive parallelism: the number of operations that can be executed at the same time is proportional to the number of molecules involved, which is of the order of 10 to the power 19 . Also, operations which involve bio-molecules are over a billion times more energy efficient with respect to electronic chips, and the information can be stored at a density of about a billion times higher than in usual electronic computers. The major applications to massively-parallel molecular computation range from novel computer architectures in conventional hardware and novel algorithmic solutions to difficult problems to self-assembling technology and intelligent nano-scale construction. The theoretical studies involve the investigation of new computational models based on paradigms coming from bio-chemistry: the complementarity of the two strands of a DNA molecule, the signaling within and between cells, or the structural organization of cells.
Funded by European Union IST FP5, 2002-2004.
Publications
Click here to see the full list of publications from the TUCS Publication Database
The latest updated publications:
Sepinoud Azimi, Diana-Elena Gratie, Bogdan Iancu, Ion Petre, Three Approaches to Quantitative Model Refinement with Applications to the Heat Shock Response. TUCS Technical Reports 1067, 2013.
Sepinoud Azimi, Bogdan Iancu, Ion Petre, Reaction Systems Models for the Heat Shock Response. TUCS Technical Reports 1075, TUCS, 2013.
Diana-Elena Gratie, Bogdan Iancu, Ion Petre, ODE Analysis of Biological Systems. TUCS Technical Reports 1072, TUCS, 2013.
Diana-Elena Gratie, Ion Petre, Quantitative Petri Nets Models for the Heat Shock Response. TUCS Technical Reports 1068, TUCS, 2013.
Giorgio Ausiello, Hendrik Jan Hoogeboom, Juhani Karhumäki, Ion Petre, Arto Salomaa (Eds.), Magic in Science. Theoretical Computer Science 429, 2012.
Jarkko Kari, Ion Petre (Eds.), Special Issue on Unconventional Computing. Natural Computing 11(4), 2012.
Ion Petre, Corrado Priami, Erik de Vink (Eds.), Special Issue on Computational Models for Cell Processes. Transactions on Computational Systems Biology Lecture Notes in Computer Science 7625, 2012.
Elena Czeizler, Eugen Czeizler, Bogdan Iancu, Ion Petre, Quantitative Model Refinement as a Solution to the Combinatorial Size Explosion of Biomodels. Electronic Notes in Theoretical Computer Science 284, 35–53, 2012.
Sepinoud Azimi, Tero Harju, Miika Langille, Ion Petre, Simple Gene Assembly as a Rewriting of Directed Overlap-Inclusion Graphs. Theoretical Computer Science 454, 30–37, 2012.
Eugen Czeizler, Andrzej Mizera, Elena Czeizler, Ralph-Johan Back, John E. Eriksson, Ion Petre, Quantitative Analysis of the Self-Assembly Strategies of Intermediate Filaments from Tetrameric Vimentins. IEEE-ACM Transactions on Computational Biology and Bioinformatics 9(3), 885–898, 2012.
Elena Czeizler, Andrzej Mizera, Ion Petre, A Boolean Approach for Disentangling the Roles of Submodules to the Global Properties of a Biomodel. Fundamenta Informaticae 116(1-4), 51–63, 2012.
Eugen Czeizler, Vladimir Rogojin, Ion Petre, The Phosphorylation of the Heat Shock Factor as a Modulator for the Heat Shock Response. IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS 9(5), 1326–1337, 2012.
Andrzej Mizera, Eugen Czeizler, Ion Petre, Self-Assembly Models of Variable Resolution. Transactions on Computational Systems Biology XIV(7625), 181–203, 2012.
Ion Petre, Sergey Verlan, Matrix Insertion–Deletion Systems. Theoretical Computer Science 456, 80–88, 2012.
Robert Brijder, Mark Daley, Tero Harju, Natasha Jonoska, Ion Petre, Grzegorz Rozenberg, Computational Nature of Gene Assembly in Ciliates. In: Grzegorz Rozenberg, Thomas Bäck, Joost N. Kok (Eds.), Handbook of Natural Computing, 1233–1280, Springer, 2012.
Andrzej Mizera, Elena Czeizler, Ion Petre, Computational methods for quantitative submodel comparison. In: Evgeny Katz (Ed.), Biomolecular Information Processing. From Logic Systems to Smart Sensors and Actuators, 323–346, Wiley-VCH Verlagsgesellschaft, 2012.
Eugen Czeizler, Vladimir Rogojin, Ion Petre, The Phosphorylation of the Heat Shock Factor as a Modulator for the Heat Shock Response. TUCS Technical Reports 1041, Turku Centre for Computer Science, 2012.