A COMPUTATIONAL APPROACH TO PREDICTIVE MODELING USING DEGREE-BASED TOPOLOGICAL DESCRIPTORS: APPLICATIONS IN STRUCTURE BOILING POINT RELATIONSHIPS OF OXO ORGANIC COMPOUNDS

In organic chemistry the prefix oxo indicates a double bond between an oxygen atom and carbon atom. The oxo group is present in ketons, estern, carboxilic acids, aldehydes and amids. Topological indices are numerical quantities associated with the molecular graphs. Some of the molecular descriptors have proven their ability in predicting physico-chemical properties of certain chemical compounds. In this paper structure boiling point relationships are studied for series of oxo organic compounds by ten molecular descriptors. The QSPR modelling for the oxo organic compounds are presented by the means of linear, quadratic, logarithmic and multilinear analysis. The multilinear analysis model produces a correlation coefficient(r) =0.950 and standard error of estimates (s) = 17.1902◦C for the set of 192 oxo organic compounds. Also, the intercorrelation between pair of molecular descriptors is studied. This is validated by Hierarchial cluster analysis. This study helps to classify useful molecular descriptors based on their predicting in structure boiling point of oxo organic compounds