Kaltenbach, Hans-Michael, Dr.

Background

Trained as a mathematician, I got my diploma degree in 2003 from the University of Hannover, Germany, with a thesis on stochastic runtime analysis of algorithms. I then joined the International Graduate School for Bioinformatics and Genome Research at the University of Bielefeld, where I got my PhD in 2007 with a thesis on efficient algorithms and statistics for protein identification using mass spectrometry. In 2007/2008 I was a postdoctoral fellow at the Institut Pasteur, Paris, France with Benno Schwikowski working on algorithms for mass spectrometry and algorithms and statistics for analysis of biological networks. In 2008, I joined Joerg Stelling's computational systems biology group at ETH Zurich, Switzerland, as a postdoctoral fellow and am a senior research assistant in this group since 2011.

Research Interest

My main research interests revolve around the development of new mathematical and computational methods to help answering biological research questions. Although my perspective is usually quite clearly influenced from mathematical stochastics, I am constantly trying to equip my toolbox with new methods (and toys) from applied and pure math, algorithms, and engineering and to gain more knowledge about wet-lab biology.

Current Projects

I am currently involved in a couple of projects that span from almost pure math to nearly (and soon really) touching a pipette. My main projects are

Structural Analysis of Qualitative Dynamics
We are working on a better understanding and extensions of several methods that allow to analyze the qualitative dynamics of biochemical reaction systems with very limited knowledge about rate laws and parameters. We mainly work with three approaches:

Chemical Reaction Network Theory
Formulated in the 1980s, CRNT allows to determine whether a given mass
action network is able to exhibit multiple steady states. We are currently working on a better understanding and exposition of this theory using linear algebra and some tools from pure algebra and graph theory.

Injectivity using species-reaction graphs
More recently, several publications suggest necessary and sufficient conditions for multistationarity of a network by analyzing the rank of its Jacobian. Due to the particular structure of the ODE systems for reaction networks, graph- and matrix-theoretic methods can used to establish whether the Jacobian is of full rank in the positive orthant of the phase-space.

Monotone Systems
Proposed by Hirsch for ODE systems and extended by Sontag to systems with inputs and outputs, monotonicity of a system implies several strong constraints on the possible dynamics of the system. For example, oscillations and chaotic behavior can be ruled out. Interestingly, monotonicity can again be established in special cases by looking at the cycle structure of associated graphs. We work on methods to decompose systems into monotone subsystems, approximations for these systems, and their applications to a coarse-grained analysis.

MAPK signaling
Together with Fabian Rudolf and Mikael Sunnaker, we are working on a new model for pheromone signaling in yeast to be able to identify and explain feedback loops in this system. From the methods side, new tools for formulating and testing alterations of models are developed. The perspective here is mainly ordinary differential equations and parameter identification problems. This project is in very close collaboration with Fabian Rudolf's yeast lab for the wet-lab part and the biological insights, swinging back and forth between in vivo and in silico.

UniCellSys
This large EU project aims at integrating several models of signal transduction, metabolism, and cell cycle to model cell proliferation and reactions to stress in S. cerevisiae. For this, we are developing new tools to integrate models and to decompose large models into modules in order to reduce the complexity of the models while still preserving the main dynamic features. Our focus is on advanced methods and tools from control theory and dynamic systems. Besides the UniCellSys consortium including modeling and biological groups, we also finished a side-project with my students Justin Feigelman and Simona Constaniescu on monotone systems decomposition. See also UniCellSys webpage.

Automatic evaluation of dynamic pathway models using contingencies
This is a project in collaboration with Marcus Krantz where we try to leverage information from biological experiments (mainly taken from the literature) to automatically evaluate the consistency of a dynamic pathway model with experimental data.

MetaNetX
With more genome-scale metabolic networks becoming available, the question of how to compare two models, how to evaluate them, and how to (semi-)automatically correct and enhance models with knowledge from databases and other models becomes more pressing. In particular, metabolism of plants is a widely open field with only few models available, but a great biological interest. Within the MetaNetX project, we are developing new metrics to evaluate stoichiometric models and develop new statistical tools for exploiting existing knowledge. This currently involves identification and development of advanced statistical methods and algorithms for their computation. For more information, see also the MetaNetX webpage .

Publications

Kaltenbach, H.-M., "A Concise Guide to Statistics", Springer, ISBN 978-3-642-23501-6, will be available from October 31, 2011. Book website at Springer. For a list of known factual errors in the book, download the ERRATA.

Kaltenbach, H.-M., Constantinescu, S., Feigelman, J. and Stelling, J., "Graph-Based Decomposition of Biochemical Reaction Networks into Monotone Subsystems", Proceedings of WABI 2011, volume 6833 of Lect. Notes Bioinformatics, pages 139-150. Springer, 2011

Kaltenbach, H.-M. and Stelling, J., "Modular Analysis of Biological Networks", in: Adv Exp Med Biol, 736:3-17, Springer, 2012

Marschall, T., Herms, I., Kaltenbach, H.-M. and Rahmann, S., "Probabilistic Arithmetic Automata and their Applications", arXiv 1011.5778, 2010, arXiv link

Kaltenbach, H.-M., Dimopoulos S. and Stelling J., "Systems analysis of cellular networks under uncertainty", FEBS Lett. 583 (24), 2009, pages 3923-3930. doi:10.1016/j.febslet.2009.10.074

Uhr, M., Kaltenbach, H.-M., Conradi, C. and Stelling, J., "Analysis of Degenerate Chemical Reaction Networks.", Proceedings of Positive Systems: Theory and Applications, LNCIS 389, pages 163-171, 2009 10.1007/978-3-642-02894-6_16

Vandenbogart, M., Li-Thiao-Th´e, S., Kaltenbach, H.-M. Zhang, R., Aittokallio, T. and Schwikowski, B. "Alignment of LC-MS images, with applications to biomarker discovery and protein identification." Proteomics, 8:650–672, 2008.

Kaltenbach, H.-M., Wilke, A. and Böcker, S. "SAMPI: Protein identification with mass spectra alignments." BMC Bioinformatics, 8:102, 2007.

Kaltenbach, H.-M., Böcker, S. and Rahmann, S. "Markov additive chains and applications to fragment statistics for peptide mass fingerprinting." In T. Ideker and V. Bafna, editors, Syst. Biol. and Comput. Proteomics Ws, volume 4532 of Lect. Notes Comput. Biol., pages 29–41. Springer, 2007

Böcker, S. and Kaltenbach, H.-M., "Mass spectra alignments and their significance." Journal of Discrete Algorithms, 5(4):714–728, 2007.

Böcker, S. and Kaltenbach. H.-M., "Mass spectra alignments and their significance." In Alberto Apostolico, Maxime Crochemore, and Kunsoo Park, editors, Combinatorial Pattern Matching, volume 3537 of Lect. Notes Comp. Sci., pages 429–441. Springer, 2005.

Kaltenbach, H.-M.: "Statistics and Algorithms for Peptide Mass Fingerprinting", PhD Thesis, Faculty of Technology, Bielefeld University, 2007

Kaltenbach, H.-M.: "Analytische und stochastische Methoden der De-Poissonisierung", Diploma Thesis, Institute for Mathematical Stochastics, Hannover University, 2003

Conference Tutorials

I gave several tutorials on structural analysis of biochemical reaction networks during the International Conference for Systems Biology (ICSB) in 2010 and 2011. Slides of these tutorials are available upon request.

ICSB 2010

• Reaction network theory for systems biology (with C. Conradi, MPI Magdeburg)

ICSB 2011

• Reaction network theory for systems biology (with C. Conradi, MPI Magdeburg)
• Injectivity and multistationarity in biochemical reaction networks (with I. Otero Muras, ETH Zurich)

Teaching

At ETH Zurich:

• Structural Analysis of Biochemical Reaction Networks (Spring 2012) Course webpage
• Fundamentals of Modern Statistics (Fall 2011) Course webpage
• Statistics for Systems Biology (Fall 2010) (the lecture script finally became the book "A Concise Guide to Statistics", see above) Course webpage
• Introduction to Practical Modeling and Simulation of Metabolic and Signaling Networks (Spring 2009) (lecture script available upon request) Course webpage

At University of Bielefeld (2003-2007):

• Statistische Methoden der Bioinformatik
• Currents in Mass Spectrometry
• Topics in Probability Theory and Statistics for Bioinformatics
• Markov Modelle in der Bioinformatik
• Advanced Algorithms
• Numerics of Stochastic Differential Equations
• Signal Processing
• Special Topics in Linear and Quadratic Programming
• Lecture Analyse von Algorithmen (analysis of algorithms, combinatorial and stochastic runtime analysis, recurrence equations, and the like)