QSAR Studies On Several Series Of Drugs Using Hqsar, CoMFA And Molecular Design

Quantitative structure-activity relationship (QSAR) functions as an importanttool in the drug design.Due to the unique structural characterization, hologram-QSARand comparative molecular field analysis (CoMFA), which have been elaborated indetail in this paper, play an increasingly important role in computer-aided drug design.Quantitative structure-activity relationship consists of two steps:(1) the calculation ofstructure descriptors;(2) the correlation of structure descriptors with activity. Inregard to the studies about structural characterization and structure-activityrelationship, HQSAR and CoMFA are important QSAR methods. HQSAR, miningstructure information from structure graphing, transforms molecular structure intoholographic structure descriptor, while CoMFA characterizes three-dimensionalstructural information from the perspective of the distribution of molecular force fieldaround the molecules, providing important insight into the interaction mode betweenthe drug and target. Due to the unique structural characterization in HQSAR andCoMFA, This paper, from the perspective of holographic structure information andforce field information, carried out QSAR studies on three types of drug, investigatingthe effect of structural information on activity, which can throw light on the idea andmethod of molecular design. The content was summarized below:1、A series of structurally related indole-5-carboxylic acids as potent inhibitorsagainst human cytosolic phospholipase A2α were subjected to hologram quantativestructure-activity relationship (HQSAR) and comparative molecular field analysis（CoMFA）. A training set containing23compounds served to establish the HQSARmodel. The best HQSAR model was generated using atoms, bond, connectivity, donorand acceptor as fragment distinction and3–6as fragment size with five componentsshowing cross-validated q2value of0.790and conventional r2value of0.961.Themodel was then employed to predict the potency of5test set compounds that wereexcluded in the training set, and a good agreement between the experimental andpredicted values was observed exhibiting the powerful predictable capability of thismodel (r2=0.605). Contribution maps indicated that the carboxylic acid moiety inpred position5of the indole scaffold and the electron-withdrawing effects contributed tothe inhibitory activity significantly. The best CoMFA model using common structurealignment approach with the bound conformation of compound26as the templateyielded the steric parameter as a major contributor (nearly80%) to the observed variations in biological activity. Based upon some key structural features derived fromthe two QSAR models, we have designed novel inhibitors of cPLA2α possessingbetter inhibitory activity.2、 A series of structurally related2,4-dioxopyrimidine-1-carboxamidederivatives as highly potent inhibitors against acid ceramidase were subjected tohologram quantative structure-activity relationship (HQSAR) analysis. A training setcontaining24compounds served to establish the HQSAR model. The best HQSARmodel was generated using atoms, bond, connectivity, donor and acceptor as fragmentdistinction and3–6as fragment size with six components showing cross-validated q2value of0.834and conventional r2value of0.965. The model was then employed topredict the potency of test set compounds that were excluded in the training set, and agood agreement between the experimental and predicted values was validatedshowing the powerful predictable capability of this model (r2pred=0.788). Atomcontribution maps indicate that the electron-withdrawing effects at position5of theuracil ring, the preferential acyl substitution at position N3and the eight-carbon alkylchain length of the N1-carboxamide function predominantly contribute to theinhibitory activity. Based upon these key structural features derived from HQSAR2Datom contribution maps, we have designed novel inhibitors of acid ceramidasepossessing better inhibitory activity.3、Cytotoxic DNA intercalating agents act as topoisomerase (topo) inhibitors.Dimeric compounds designed as potential bis-intercalating agents show significantlyenhanced potencies in cell culture and in many cases broad-spectrum in vivo activityas well. To get more information about the inhibition mechanism and construct thebinding model between the ligand and the plausible receptor, comparative molecularfield analysis (CoMFA) was performed to explore the structure-activity relationshipof substituted bis[(acridine-4-carboxamides) propyl]methylamines. The model had anon-cross validated coefficient (r2) value of0.917, and a cross-validated (q2) value of0.586, which indicates good statistic stability and predictive power. Based on theanalysis of CoMFA model thus derived, we have designed some novel molecules withhigher activities. Using both HQSAR and CoMFA methods, which characterizes the molecularstructure from the perspective of molecular hologram and molecular force fieldrespectively, a series of QSAR studies were conducted in this paper. Through theanalysis and discussion of the established QSAR models, the paper laid emphasis onthe effect of molecular structure information on drug activity, especially theholographic structure information and force field information, on the basis of which,molecular design was innovatively achieved. Three, six and three new molecules withpossibly enhanced drug activity, respectively corresponding to the three types of druginvestigated in this paper, have been designed, which has been rarely reported in theQSAR-related literature.