In the field of physiological studies there is an increasing number of molecular level pathway databases as well as detailed information on gene regulation mechanisms, protein-protein networks, gene-disease networks, genetic and epigenetic variations, thanks to the recent evolution in measuring technology. Gene regulation networks are of key importance for this knowledge but processing and interpreting them proves difficult.
My study’s primary focus is on the easier processing of gene regulation networks while interpreting them as probability networks. I look at how the pathways downloaded from the various publicly accessible pathway databases can be represented as Bayesian networks. On the basis of the Kyoto Encyclopedia of Genes and Genomes (KEGG) database I’ll present a few ways to process pathways, the different models for gene regulation pathways, the possibilities and limitations of transformation into probability networks. The model is implemented in the BayesCube software developed by the Department of Measurement and Information Systems. I also examine the possibilities behind inference and subgraph extraction in the resulting probability networks. At last I cover the technologies used in this field of bioinformatics and this study’s future development.