Martin Pelikán

In this paper, an algorithm based on the concepts of genetic algorithms that uses an estimation of a probability distribution of promising solutions in order to generate new candidate solutions is proposed. To estimate the distribution, techniques for modeling multivariate data by Bayesian networks are used. The proposed algorithm identifies, reproduces and mixes building blocks up to a specified order. It is independent of the ordering of the variables in the strings representing the solutions. Moreover, prior information about the problem can be incorporated into the algorithm. However, prior information is not essential. Preliminary experiments show that the BOA outperforms the simple genetic algorithm even on decomposable functions with tight building blocks as a problem size grows. 1 INTRODUCTION Recently, there has been a growing interest in optimization methods that explicitly model the good solutions found so far and use the constructed model to guide the fu...

Flexibility between domains of proteins is often critical for function. These motions and proteins with large scale flexibility in general are often not readily amenable to conventional structural analysis such as X-ray crystallography, nuclear magnetic resonance spectroscopy (NMR) or electron microscopy. A common evolution of a crystallography project, once a high resolution structure has been determined, is to postulate possible sights of flexibility. Here we describe an analysis tool using relatively inexpensive small angle X-ray scattering (SAXS) measurements to identify flexibility and validate a constructed minimal ensemble of models, which represent highly populated conformations in solution. The resolution of these results is sufficient to address the questions being asked: what kinds of conformations do the domains sample in solution? In our rigid body modeling strategy BILBOMD, molecular dynamics (MD) simulations are used to explore conformational space. A common strategy is to perform the MD simulation on the domains connections at very high temperature, where the additional kinetic energy prevents the molecule from becoming trapped in a local minimum. The MD simulations provide an ensemble of molecular models from which a SAXS curve is calculated and compared to the experimental curve. A genetic algorithm is used to identify the minimal ensemble (minimal ensemble search, MES) required to best fit the experimental data. We demonstrate the use of MES in several model and in four experimental examples.