Probabilistic metabolic pathways construction and analysis tool

View all posters

Jurijs Meitalovs

Latvia University of Agriculture, Latvia

Computational automated methods for biochemical pathways constructions are becoming more popular each year. The availability of data of genomes and biochemical reactions allows usage of computers to build probabilistic metabolic routes from an initial compound to a product compound that can be implemented into a living organism. Various computational methods and tools were developed in last decades for metabolic network construction, analysis and optimization. In this work we present the software tool that can be used to predict possible minimal reactions list and its corresponded genes to build novel pathways. It can be used to reach necessary functionality of genetic modification and implement it into selected microorganism chassis like e.coli or yeast. This tool is developed in Matlab environment. To construct a pathway from the input compound to the product, the probabilistic pathway construction algorithm is used. The algorithm is based on data obtained from KEGG database. The algorithm constructs the biochemical network in levels, starting from input substrate it adds reactions into chassis organism model (biochemical model of microorganism in SBML format) and checks, is the source metabolite connected with the product or not. Pathways with smaller number of reactions get an advantage. For these reactions next analysis steps can be performed. The tool uses productivity of Cobra toolbox for biochemical model manipulation (adding and removing reactions and metabolites) and constructed pathways analysis – FBA, FVA, EMA etc.