Quantitative structure-property relationships for the prediction of vapor pressures of organic compounds from molecular structures

A quantitative structure-property relationship (QSPR) is developed to relat e the molecular structures of 420 diverse organic compounds to their vapor pressures at 25 degrees C expressed as log(vp), where vp is in pascals. The log(vp) values range over 8 orders of magnitude from -1.34 to 6.68 log uni ts. The compounds are encoded with topological, electronic, geometrical, an d hybrid descriptors. Statistical and computational neural network (CNN) mo dels are built using subsets of the descriptors chosen by simulated anneali ng and genetic algorithm feature selection routines. An 8-descriptor CNN mo del, which contains only topological descriptors, is presented which has a root-mean-square (rms) error of 0.37 log unit for a 65-member external pred iction set. A 10-descriptor CNN model containing a larger selection of desc riptor types gives an improved rms error of 0.33 log unit for the external prediction set.