Abstract

This paper aimed to present a methodology for the geometry optimization of molecules in order to analyze structure-activity/property relationships. The main objective was to reduce the probability of carrying out misleading analyses while preparing the molecules for modeling. An algorithm that used HyperChem 8.0 software was developed and implemented in a series of PHP programs. A sample of 319 drug-like compounds was used to test the implemented algorithm. The program executed the geometry optimization protocols efficiently and quickly on both small and large sample sizes.  The results obtained in the investigated sample of 319 compounds showed that our programs performed well on the investigated compound collection thus increasing the reliability of compound preparation for quantitative structure-activity relationships analysis.

Keywords

Computer Assisted/Automated System, Geometry Optimization, in silico Modeling, Accuracy