Markush Toolkit

Barnard Chemical Information
Reviewed by Susan Boyd
CompChem Solutions

Barnard Chemical Information (BCI) has supplied chemoinformatics applications for over 15 years, but has recently launched a series of programming toolkits for users to build their own, custom applications in-house. The latest offering in this line is the Markush Toolkit, which offers the ability to enumerate 2D chemical structures extremely rapidly - at least 50 000 structures per second is a conservative estimate for enumeration with property calculation - performs full structure (including overlap) and substructure searches, and computes various 2D properties of the compounds, including Lipinski properties, BCI fingerprints or the user’s own custom properties, in equally impressive timescales. These properties are computed additively and hence can be output - and subsequently used as filters - prior to any data dump of enumerated structures. This opens the way to data mining of very large datasets indeed, although sets selected for synthesis can be as large or as focused as required by the user.

Use of the toolkits does require some programming knowledge, but a reasonable programmer with some knowledge of chemistry should find the toolkits easy to use, and intuitive to learn. BCI supplies a number of example source programs and full documentation to get new users started.

The toolkits come with C/C++, Java, Visual Basic, Perl or Python wrappers, and are supported on Windows, Linux, Irix and Solaris platforms.

For further information on the Markush Toolkit and associated BCI toolkits (including clustering and diversity applications) contact BCI.