Scaffold fragmentation and substructure hopping reveal potential, robustness, and limits of computer-aided pattern analysis (C@PA).

Computer-aided pattern analysis (C@PA) was recently presented as a powerful tool to predict multitarget ABC transporter inhibitors. The backbone of this computational methodology was the statistical analysis of frequently occurring molecular features amongst a fixed set of reported small-molecules that had been evaluated toward ABCB1, ABCC1, and ABCG2. As a result, negative and positive patterns were elucidated, and secondary positive substructures could be suggested that complemented the multitarget fingerprints. Elevating C@PA to a non-statistical and exploratory level, the concluded secondary positive patterns were extended with potential positive substructures to improve C@PA's prediction capabilities and to explore its robustness. A small-set compound library of known ABCC1 inhibitors with a known hit rate for triple ABCB1, ABCC1, and ABCG2 inhibition was taken to virtually screen for the extended positive patterns. In total, 846 potential broad-spectrum ABCB1, ABCC1, and ABCG2 inhibitors resulted, from which 10 have been purchased and biologically evaluated. Our approach revealed 4 novel multitarget ABCB1, ABCC1, and ABCG2 inhibitors with a biological hit rate of 40%, but with a slightly lower inhibitory power than derived from the original C@PA. This is the very first report about discovering novel broad-spectrum inhibitors against the most prominent ABC transporters by improving C@PA.
Competing Interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(© 2021 The Author(s).)