Targeting WEE1 kinase: an integrated machine learning-cheminformatics framework for ultra-large-scale virtual screening and novel inhibitor discovery.

Authors

Rajesh Muthuraj, Manasa Pacharla, Nehal Arvind Kumar, Dhanushya Gopal, Mohit Agrawal, Angelin Preetha Sathiyanathan, Prem Kumar, Rosemary Edwin, Jaikanth Chandrasekaran

Published in

Journal of computer-aided molecular design. Volume 40. Issue 1. Aug 14, 2026. Epub Aug 14, 2026.

Abstract

WEE1 kinase, a critical regulator of the G2/M checkpoint, represents a validated therapeutic target in tumors harboring defects in DNA damage response (DDR) pathways. Although clinical inhibitors such as adavosertib have demonstrated therapeutic potential, challenges, including selectivity constraints, dose-limiting toxicities, and emerging resistance, highlight the need to expand the structural diversity of WEE1-targeting chemotypes. Here, we report a scalable, machine-learning-integrated virtual screening framework designed to explore ultra-large chemical space spanning an input search space of approximately 884 million compounds from ZINC20 and 199,854 purchasable compounds from the SPECS database. Molecular representations using ECFP4 fingerprints combined with UMAP-based dimensionality reduction and K-means clustering enabled diversity-guided prioritization across distinct regions of chemical space. Multi-stage structure-based computational evaluation, including pharmacophore modelling, molecular docking, MM/GBSA rescoring, and 200-ns molecular dynamics simulations, yielded 18 high-confidence candidates, from which three structurally novel scaffolds were selected for detailed analysis. One ZINC-derived and two SPECS-derived compounds demonstrated predicted stable binding modes involving key WEE1 active-site residues and favourable estimated developability profiles. Critically, in silico selectivity profiling against the off-target PLK1 revealed structurally grounded differential binding, providing a computational basis for selectivity. Preliminary in vitro evaluation of one SPECS-derived compound (AJ-292/13095349) demonstrated antiproliferative activity in triple-negative breast cancer (TNBC) models. Collectively, this study establishes an efficient computational hit identification framework for ultra-large screening and reports structurally distinct starting points for WEE1-targeted oncology drug discovery.

PMID:
42599595
Bibliographic data and abstract were imported from PubMed on 15 Aug 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.