Protein family tunalbe scoring functions; Improving enrichment with
eHiTS.

Darryl Reid [a], Zsolt Zsoldos [a] , Aniko Simon [a], A. P. Johnson [b]

[a] SimBioSys Inc., 135 Queens Plate Dr, Unit 520, Toronto, Ont., M9W 6V1, Canada.
[b] ICAMS, University of Leeds, Leeds, LS2 9JT, UK

The key to any docking program is the ability to score and rank potential ligands to enrich the top portion of the dataset with actives.  Most docking programs offer a single generic scoring function intended to represent all proteins equally well.  However, it is well known that each scoring function "favors" particular families or types of proteins over others.  Here we present results illustrating the latest approach to this problem by the eHiTS docking program.  By introducing protein-family tunable scoring functions, eHiTS is able to adjust to different protein families by "perceiving" which scoring function to use with each protein it is given.  A set of enrichment studies will be presented to illustrate the effect of trained scoring functions. More information: http://www.simbiosys.com/