Are the Chemical Structures in Your QSAR Correct?

TLDR

QSARs predict diverse endpoints using many descriptors, but all rely on the assumption that the underlying chemical structures are correct. The study examines the validity of structural correctness in six databases and uses a case study on octanol/water partition coefficient predictions to illustrate the impact of errors. The authors used molecular fingerprinting to assess error rates in six databases and built QSARs with both correct structures and a 3.4%‑error database to evaluate prediction differences. The databases exhibited 0.1–3.4% error rates, and the case study demonstrated that minor structural mistakes can markedly reduce QSAR prediction accuracy.

Abstract

Abstract Quantitative structure–activity relationships (QSARs) are used to predict many different endpoints, utilize hundreds, and even thousands of different parameters (or descriptors), and are created using a variety of approaches. The one thing they all have in common is the assumption that the chemical structures used are correct. This research investigates this assumption by examining six public and private databases that contain structural information for chemicals. Molecular fingerprinting techniques are used to determine the error rates for structures in each of the databases. It was observed that the databases had error rates ranging from 0.1 to 3.4%. A case study to predict the n ‐octanol/water partition coefficient was also investigated to highlight the effects of these errors in the predictions of QSARs. In this case study, QSARs were developed using both (i) all correct structures and (ii) structures from a database with an error rate of 3.4%. This case study showed how slight errors in chemical structures, such as misplacing a Cl atom or swapping hydroxy and methoxy functional groups on a multiple ring structure, can result in significant differences in the accuracy of the prediction for those chemicals.