Application of PARM to constructing and comparing 5-HT1A and a1 receptor models

A combination of genetic algorithm models and a cross validation technique were used to produce at. level pseudo-receptor models starting from a set of known ligands. This was based on Walters' GERM (genetic evolved receptor model) and PARM (pseudo at. receptor model). These putative pseudo-receptor models can be used to predict the bioactivity of virtual mols. by aligning these mols. with the training set mols., computing the interaction energy between each mol. and interpolating the computed interaction energy in the quant. structure-activity relationship (QSAR) regression equation to obtain a predicted bioactivity, thereby reducing the trial and error procedure in the synthesis of new chem. entities.