Search

The apparatus of Generalized Nets (GN) is applied here to a description of a selection operator, which is one of the basic genetic algorithm operators. The GN model presented here describes roulette wheel selection. The resulting GN model could be considered as a separate module, but it can also be accumulated into a GN model to describe a whole genetic algorithm.

Three software packages based on the common platform of AMMOS (Automated Molecular Mechanics Optimization tool for in silico Screening) for assisting virtual ligand screening purposes have been recently developed. DG-AMMOS allows generation of 3D conformations of small molecules using distance geometry and molecular mechanics optimization. AMMOS_SmallMol is a package for structural refinement of compound collections that can be used prior to docking experiments. AMMOS_ProtLig is a package for energy minimization of protein-ligand complexes. It performs an automatic procedure for molecular mechanics minimization at different levels of flexibility - from rigid to fully flexible structures of both the ligand and the receptor. The packages have been tested on small molecules with a high structural diversity and proteins binding sites of completely different geometries and physicochemical properties. The platform is developed as an open source software and can be used in a broad range of in silico drug design studies.

In this work, a modified multi-population genetic algorithm is developed for the purpose of parameter identification of fermentation process model. Modified multi-population genetic algorithm is similar to the multi-population one and its development is instigated by modified genetic algorithm, similar to simple one. A comparison of four types of genetic algorithms, namely simple, modified, multipopulation and modified multi-population is presented for parameter identification of a fed-batch cultivation of Saccharomyces cerevisiae

Free software for entering and documenting data EpiData is here used for design of a system for functional states recognition during a fermentation process. The identification of the current process state is based on the predetermined rules, rendering specific metabolic mechanisms. Developed system is further applied for a fed-batch cultivation of Saccharomyces cerevisiae.

Free software for entering and documenting data EpiData is here used for design of an advisory system for on-line control of a fermentation process. Based on the preliminary developed system for functional state recognition, presented here system will advise the user which new functional state can be reached and what kind of control actions have to be taken. New-designed system appears as an expert system and comprises knowledge of well-trained operators of cultivation processes. Developed advisory system is further applied for a fed-batch cultivation of Saccharomyces cerevisiae.