Detour Cum Distance Matrix Based Topological Descriptors for QSAR/QSPR Part-I: Development and Evaluation

Abstract

The structure of the compound depends on connectivity of its constituent atoms. Topological descriptors based on connectivity can reveal the role of structural and substructural information of molecules in estimating biological activity. The present study involves conceptualization of six detour cum distance matrix based topological descriptors (TDs) termed as relative distance sum descriptors and relative distance product descriptors (denoted by S1R , S2R , S3R , S4R , S1RP and S2RP ) as well as their topochemical versions (denoted by S1Rc, S2Rc, S3Rc S4Rc, S1RPc and S2RPc ). The proposed descriptors have been specifically designed to take care of the molecules containing cyclic substituents. An inhouse computer program was utilized to compute the values of the proposed TDs for all possible four, five and six membered hydrogen depleted structures. Proposed TDs were assessed for similarity analysis, intercorrelation, degeneracy, discriminating power, and sensitivity towards branching apart from relative position of substituent in molecules containing cyclic moieties. The said TDs exhibited negligible degeneracy, high sensitivity towards branching/relative position of substituent(s) in cyclic structures amalgamated with exceptionally high discriminating power. The above mentioned attributes provide proposed descriptors a huge potential for use in characterization of molecules, lead identification/ optimization, combinatorial library prototype, dissimilarity/similarity studies and (Q)SAR/QSPR/QSTR/QSPkR studies so as to facilitate drug design. This has been further supported by application of the proposed descriptors on the real dataset in the part-II of the manuscript.