Prediction Of Properties Of Macromolecular Organic Compounds (Drug) With The Positional Distributive Contribution Theory

Physicochemical properties of organic compounds are important for predicting their behavior in chemical engineering design and experiments. Because the experimental measurement of physicochemical properties is expensive and time-consuming, the alternative is to develop models from which the required properties with the desired accuracy could be obtained.Based on position group contribution additivity, a new group contribution method has been developed and applied for the prediction of physicochemical properties of organic compounds. And this method has been used successfully for prediction of vapor pressure, critical temperature, critical pressure, boiling point etc. Here, this new method is named as Positional Distributive Contribution Method (PDCM). Owing to the utilization of the PDCM, we define it as function. Position factor could distinguish between most isomers including cis-or trans-or Z-and E-structures of organic compounds for their properties.A function of Positional Distributive Contribution was performed on aryl hydrocarbon (Ah) receptor activity and mutagenic potency (described as pEC50) of127heterocyclic compounds and anti-HIV activity of89TIBO derivatives in this article. The objective of this work were to determine whether a more general structure-pEC50relationship and structure-IC50relationship based solely on one topological index, could be developed through the systematic QSAR(quantitative structure activity relationship) approach. The descriptors, so-called topological matrix, used in our study are innovative and concise compared to the combination of topological, structural, physicochemical. electronic and spatial descriptors.The results indicate that our topological index provides very satisfactory results. The overall average absolute difference (AAD) and the relative errors for pEC50predictions of32dibenzofurans derivatives are found to be0.40and7.8%, respectively. While with TS+TC ridge regression model, the AAD for pEC50prediction is0.45and the mean absolute relative derivation is8.7%. Obviously our method is better than TS+TC ridge regression model. In the same way, the data of pEC50predictions of95aromatic and heteroaromatic amines and IC50activity of89TIBO derivatives is superior to others. Comparing with others method, our method performs better both in accuracy and generality.