Prediction of Hydroxyl Radical Rate Constants from Molecular Structure

Quantitative structure−property relationships are developed using multiple linear regression and computational neural networks (CNNs). Structure-based descriptors are used to numerically encode molecular features that can be used to form models describing reaction rates with hydroxyl radicals. For a set of 57 unsaturated hydrocarbons, a 5−2−1 CNN was developed that produced a root-mean-square (rms) error of 0.0638 log units for the training set and 0.0657 log units for an external prediction set. The residual sum of squares for all 57 compounds was 0.234 log units, which compares very favorably with existing methodologies. Additionally, a 10−7−1 CNN was built to predict hydroxyl radical rate constants for a diverse set of 312 compounds. The training set rms error was 0.229 log units, and the rms error for the external prediction set was 0.254 log units. This model demonstrates the ability to provide accurate predictions over a wide range of functionalities.