VALIDATION OF EGSITE2, A MIXED-INTEGER PROGRAM FOR DEDUCING OBJECTIVESITE MODELS FROM EXPERIMENTAL BINDING DATA

EGSITE2 represents a substantial advance in a long series of methods f or calculating receptor site models given only specific binding data. Compared to our most recently reported technique, EGSITE [Schnitker et al. J. Comput.-Aided Mel. Des. 1997, 11, 93-110] the user no longer h as to simplify the structures of the molecules in the training set by clustering the atoms into a few superatoms. The only remaining source of subjectivity is the user's choice of compounds for the training set , which can be surprisingly few in number. Then EGSITE2 automatically produces typically several different models that explain the observed binding without outliers. The models are remarkably simple but have su bstantial predictive power for any sort of test compound, with an esti mation of the uncertainty of the prediction. Validation of the method is reported for four standard test cases: triazines and pyrimidines bi nding to dihydrofolate reductase, steroids binding to corticosteroid-b inding globulin and to testosterone-binding globulin, and peptides bin ding to angiotensin-converting enzyme.