Construction of a chemical structure searching system using only open source softwares

Abstract

Chemical databases which accumulate information such as chemical and 3D structures, their properties of compounds are one of the core elements in cheminfomatics. When chemists construct their own database of compounds handled in their laboratories, there are solutions such as Oracle database + MDL Direct, THOR-Merlin system and so on. They are too expensive to use small database in laboratory although they are very powerful to construct desired database. In the present study, we tried to build a database system which can search and find compounds with similar chemical structures using only opensource softwares. The present system is consist of Linux(CentOS 4.4) as OS, PostgreSQL(7.4)+pgchem::tigress(7.0) as RDBMS, Apache(1.3.22)+PHP(5.0) for the Web server, and mol2ps(0.1d)+ImageMagick(6.3) for displaying chemical structures. Pgchem::tigress is an opensource software developed by Ernst-Georg Schmid et al. The program can find chemical structures on PostgreSQL. We made an exemple database including 4669 structural data registered in ChemBank BioActivities. It is possible to search both the same structure and similar substructures in this database.