POEM

Parameter Optimisation and Estimation Methods

The goal of this project is to provide comprehensive and flexible software packages for parameter estimation in systems biology, chemical reaction dynamics, nonlinear mechanics, and other fields. Special emphasis is put on an easy-to-use graphical user interface (GUI) and state-of-the-art numerical methods to solve highly nonlinear and ill-posed inverse problems. The software BioPARKIN has especially been designed for modeling , parameter identification and simulation in systems biology. The GUI uses an intuitive network editor and adheres to systems biology standards such as the SBML file format. The underlying numerical framework uses affine invariances and adaptive algorithms as given in [1]. The package will be available as source code in Python (GUI part) and C++ (numerical part).

BioPARKIN

Based on preliminary and current work at ZIB [2, 3], BioPARKIN is focused on parameter identification problems in Systems Biology. The handling of models containing species, parameters, and mechanisms is made as intuitive and simple as possible. In particular, the software facilitates rapid prototyping of new models or improving existing ones. These modeling steps are backed by providing visual feedback wherever possible.

By making use of Python and Qt [4], BioPARKIN will be available on multiple platforms including Windows, Mac OS, and Linux.

Methodology

The numerical basis of the software package was established by the theoretical work in [1], using ideas of affine invariant Newton methods.

Additionally, parameter scaling and rank deficiency, possibly occurring within single Newton steps, are thoroughly dealt with. As a result, the routines implemented for parameter identification are reliable and robust. Moreover, globalization techniques are applied to obtain maximal regions of convergence.

The extendability of the software package is ensured by a clean object-oriented design using the C++ programming language. In addition, run time efficiency is a central aspect for the implemention of the software package.

Case Study

A cooperation with aokin AG aims to develop a method for identifying the properties of antibody bonds using kinetic fluorescence polarisation. A kinetic reaction model is being developed and simulated at the Zuse Institute in order to reduce the number of measurements and the amount of material needed in aokin's laboratories. By comparing measurements with the model's predictions, as many unknown parameters as possible are to be identified.

The algorithms implemented as part of the POEM project enable the simulation of this model.

References

[ 1 ] P. Deuflhard, Newton Methods, Springer, 2004.
[ 2 ] U. Nowak, and P. Deuflhard, Numerical Identification of Selected Rate Constants in Large Chemical Reaction Systems, Appl. Num. Math. 1 (1985), 59 -- 75.
[ 3 ] T. Dierkes, M. Wade, U. Nowak, S. Röblitz: BioPARKIN: Biology-related Parameter Identification in Large Kinetic Networks, ZIB-report 11-15, 2011.
[ 4 ] http://qt.nokia.com

back

Projektsite

Research