PBSoft is a research-based innovative IT company organized by the group of employees and PhD students from the Computer Proteomics Laboratory of the ICG SB RAS. It is a small innovation enterprise specialized in development of software for text-mining and knowledge bases in the field of systems biology. PDSoft focuses on development of:
• Computer software for automated extraction of knowledge from the texts of scientific publication and databases (text-mining, database-mining);
• High quality knowledge bases for the biomedicine, biotechnology, nanobiotechnology and pharmacology;
• Computer tools for the automated reconstruction of semantic networks for molecular-genetic interactions, regulations and pathways in cell.
The PBSoft company is the winner in the "Perspective business" nomination in the First Siberian Venture Fairs, 2007, Russia. PBSoft jointly with ICG performed work on the following state contracts: (1) «Cell-Textmining: Development of the methods and software tools for extraction and integration of knowledge on molecular interactions in cell from factual and textual databases»; (2) «Development of databases in the field of nanobiotechnologies as informational infrastructure elements of the nanoindustry»; (3) «Development of the web-based system for experts, assigned for identification of interconnected proteins, revealed with post-genomics methods». PBSoft developed the software packages ANDCell and ANDVisio which allow to reconstruct the associative networks on the basis of semantic analysis of publications.
PBSoft has also developed the PDBSite database for the spatial structures of protein functional sites, containing data on about 100 000 sites from various proteins (sites of posttranslational modification, enzymatic activity, ligand binding, protein-protein and protein-RNA/DNA interactions). PDBSiteScan, a program for the recognition of sites in the spatial protein structures (http://wwwmgs.bionet.nsc.ru/mgs/systems/fastprot/pdbsitescan.html), was developed and integrated with the PDBSite database. Both the PDBSite and the PDBSiteScan systems can be used to conduct studies concerning the protein functional annotation and molecular interaction reconstruction. The WebProAnalyst (http://wwwmgs.bionet.nsc.ru/mgs/programs/panalyst/) tool which allows to reveal the correlations between protein activity and amino acid physicochemical characteristics in queried sequences has also been developed. This method can be used for analysis of quantitative structure-activity relationships in proteins. The ProtStability tool for prediction of mutation effect on protein thermodynamic stability was developed. A computer system for automated data extraction from PubMed and biological databases was developed. This computational system (ANDCell) allows user to extract information on molecular genetic interaction, gene regulation events, catalytic process, genetic polymorphisms and their associations with diseases.
The company actively participates in international projects of various scales.