TOWARD SIMILARITY MEASURES FOR MACROMOLECULAR BODIES - MEDLA TEST CALCULATIONS FOR SUBSTITUTED BENZENE SYSTEMS

A new method is proposed for the evaluation of numerical similarity me asures for large molecules, defined in terms of their electron density (ED) distributions. The technique is based on the Molecular Electron Density Lego Assembler (MEDLA) approach, proposed earlier for the gene ration of ab initio quality electron densities for proteins and other macromolecules. The reliability of the approach is tested using a fami ly of 13 substituted aromatic systems for which both standard ab initi o electron density computations and the MEDLA technique are applicable . These tests also provide additional examples for evaluating the accu racy of the MEDLA technique. Electron densities for a series of 13 sub stituted benzenes were calculated using the standard ab initio method with STO-3G, 3-21G, and 6-31G* basis sets as well as the MEDLA approa ch with a 6-31G* database of electron density fragments. For each typ e of calculation, pairwise similarity measures of these compounds were calculated using a point-by-point numerical comparison of the EDs. Fr om these results, 2D similarity maps were constructed, serving as an a id for quick visual comparisons for the entire molecular family. The M EDLA approach is shown to give virtually equivalent numerical similari ty measures and similarity maps as the standard ab initio method using a 6-31G* basis set. By contrast, significant differences are found b etween the standard ab initio 6-31G* results and the standard ab init io results obtained with smaller STO-SG and 3-21G basis sets. These te sts indicate that the MEDLA-based similarity measures faithfully mimic the actual, standard ab initio 6-31G* similarity measures, suggestin g the MEDLA method as a reliable technique to assess the shape similar ities of proteins and other macromolecules. The speed of the MEDLA com putations allows rapid, pairwise comparisons of the actual EDs for a s eries of molecules, requiring no more computer time than other simplif ied, less detailed representations of molecular shape. The MEDLA metho d also reduces the need to store large volumes of numerical density da ta on disk, as these densities can be quickly recomputed when needed. For these reasons, the proposed MEDLA similarity analysis technique is likely to become a useful tool in computational drug design. (C) 1995 by John Wiley & Sons, Inc.