Simultaneous estimation of stability constants of Mg, Ba, Ca, and Sr complexes using a small subset of molecular descriptors

Abstract

This work develops a model for the prediction of stability constants of Mg, Ba, Ca, and Sr complexes, simultaneously. Quantitative structure–property relationship models are obtained by selecting descriptors from diverse functional groups, topological, and quantum chemical descriptors. Linear and non-linear methods such as multiple linear regression, partial least squares, and artificial neural networks with Levenberg–Marquardt (ANN-LM) back propagation algorithm were used to investigate the correlation between stability constant (log K) and molecular descriptors. In this work, a suitable small subset of descriptors has been selected that is able to predict the stability constants of Mg, Ba, Ca, and Sr complexes, simultaneously. Finally, a 5-4-4 ANN-LM model was designed for predicting the stability constants of complexes with high predictive ability.

Keywords:

• Stability constant
• Alkaline earth metal complexes
• Quantitative structure–property relationship
• Molecular descriptors

Acknowledgment

Support of this work by the council research of Vali-e-Asr University is acknowledged.