Discovery of Nanomolar CDK2 Inhibitors Using Boltz-2
9 affiliations
- 1Vector Institute for Artificial Intelligence, W1140-108 College St., Toronto, ON, Canada, M5G 0C6
- 2Department of Chemistry, University of Toronto, 80 St. George St., Toronto, ON, Canada, M5S 3H6
- 3Acceleration Consortium, 700 University Ave., Toronto, ON, Canada, M7A 2S4
- 4Department of Computer Science, University of Toronto, 40 St. George St., Toronto, ON, Canada, M5S 2E4
- 5Canadian Institute for Advanced Research (CIFAR), 661 University Ave., Toronto, ON, Canada, M5G 1M1
- 6Department of Chemical Engineering & Applied Chemistry, University of Toronto, 200 College St., Toronto, ON, M5S 3E5
- 7Department of Materials Science & Engineering, University of Toronto, 184 College St., Toronto, ON, M5S 3E4
- 8Institute of Medical Science, University of Toronto, 1 King’s College Circle, Toronto, ON, M5S 1A8
- 9NVIDIA, 431 King St. W #6th, Toronto, ON, M5V 1K4
Abstract
The computational design of small-molecule ligands for pharmaceutical targets has historically been hindered by a strict trade-off between screening throughput and predictive accuracy. While next-generation cofolding affinity models like Boltz-2 hold immense promise for bridging this gap, their utility in prospective medicinal chemistry campaigns remains largely unexplored. To date, virtual screens leveraging these models have been overwhelmingly restricted to standard commercial catalogs, leaving their capacity to navigate novel, synthetically tailored chemical spaces unevaluated. Here, we present a novel hit discovery platform that integrates Boltz-2’s affinity module with a custom screening space of stereochemically complex tetrahydroquinolines, accessed efficiently via a single-step multicomponent approach. Boltz-2 successfully guided both the initial high-throughput virtual screening of these targeted scaffolds and subsequent hit optimization, rapidly yielding nanomolar inhibitors of cyclin-dependent kinase 2 (CDK2).
Related
- IMPECCABLE: A Digital Twin Drug Discovery Workflow Comprising AI, Physics and Experiment & the Creation of Nanomolar-Potent Inhibitors
- Reliability of AI Methods in Drug Discovery: Evaluation of Boltz-2 for Structure and Binding Affinity Prediction
- ADMET property prediction and toxicity profiling of CDK2 inhibitors: a practical cheminformatics tutorial using ADMET-AI 2.0 and RDKit
- Discovery of selective, metabolically stable pyrazole-based FLT3 inhibitors for the treatment of acute myeloid leukemia
- Machine-learning-guided discovery of selective CK2 inhibitors with reduced CDK off-target activity
BibTeX
@misc{papidocha2026discovery,
title = {{Discovery of Nanomolar CDK2 Inhibitors Using Boltz-2}},
author = {Sven M. Papidocha and Yoshihiro Ogura and Jiahe Ren and Mikhail Saykin and Stuart Green and Varinia Bernales and Alán Aspuru-Guzik},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15010755/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010755/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Discovery of Nanomolar CDK2 Inhibitors Using Boltz-2
AU - Sven M. Papidocha
AU - Yoshihiro Ogura
AU - Jiahe Ren
AU - Mikhail Saykin
AU - Stuart Green
AU - Varinia Bernales
AU - Alán Aspuru-Guzik
PY - 2026
DA - 2026/10/21
PB - ChemRxiv
DO - 10.26434/chemrxiv.15010755/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010755/v1
AB - The computational design of small-molecule ligands for pharmaceutical targets has historically been hindered by a strict trade-off between screening throughput and predictive accuracy. While next-generation cofolding affinity models like Boltz-2 hold immense promise for bridging this gap, their utility in prospective medicinal chemistry campaigns remains largely unexplored. To date, virtual screens leveraging these models have been overwhelmingly restricted to standard commercial catalogs, leaving their capacity to navigate novel, synthetically tailored chemical spaces unevaluated. Here, we present a novel hit discovery platform that integrates Boltz-2’s affinity module with a custom screening space of stereochemically complex tetrahydroquinolines, accessed efficiently via a single-step multicomponent approach. Boltz-2 successfully guided both the initial high-throughput virtual screening of these targeted scaffolds and subsequent hit optimization, rapidly yielding nanomolar inhibitors of cyclin-dependent kinase 2 (CDK2).
ER -