The complexation of metal ions with various organic ligands in water: Prediction of stability constants by QSPR ensemble modelling

Quantitative structure–property relationship modelling of the stability constants (log K) for the 1:1 (M:L) complexes of metal ions (M = Li+, Na+, K+, Be2+, Al3+, Ga3+, In3+, VO2+, Fe3+, Th4+, NpO2+, Am3+) with structurally diverse organic ligands in aqueous solution was performed using ensemble multiple linear regression (eMLR) analysis, support vector machines, associative neural networks and substructural molecular fragments’ descriptors. The models were validated with cross-validation procedures and with complementary external test set. For eMLR in the 5-fold cross-validation, root-mean squared error of log K varies from 0.49 (Li+) to 2.30 (In3+), and it is comparable with the systematic errors in experimental data. Designed predictor for end users implements consensus models together with the estimation of their applicability domain.