Machine Learning and Virtual Screening Methods to Discover Potential Cyclin-Dependent Kinase 2 (CDK2) Inhibitors

Area of Science:

• Computational chemistry and cheminformatics
• Drug discovery and development
• Molecular biology and oncology

Background:

• Cyclin-dependent kinase 2 (CDK2) is crucial for cell cycle regulation and a significant target for cancer therapies.
• Developing novel CDK2 inhibitors is essential for advancing cancer treatment strategies.

Purpose of the Study:

• To identify novel inhibitors of cyclin-dependent kinase 2 (CDK2) using an integrated computational approach.
• To leverage machine learning and structure-based methods for predicting and validating potential drug candidates.

Main Methods:

• A computational pipeline integrating Lipinski's Rule of Five, machine learning (ML) prediction, molecular docking, and molecular dynamics simulations (MDs) was employed.
• A random forest model was trained on a CDK2 inhibitor dataset, achieving high performance (AUC-ROC 0.90, accuracy 0.84).
• Potential inhibitors were docked against CDK2 (PDB ID: 2FVD) and further analyzed using 100 ns MD simulations.

Main Results:

• The study predicted 187 compounds as active CDK2 inhibitors.
• Two compounds, STOCK4S-00019 and STOCK4S-00025, exhibited docking scores comparable to the reference ligand.
• Molecular dynamics simulations confirmed the stable binding and favorable interactions of these two compounds with CDK2.

Conclusions:

• STOCK4S-00019 (hit1) and STOCK4S-00025 (hit2) show significant potential as novel CDK2 inhibitors.
• These findings validate the efficacy of integrated computational strategies in identifying anticancer drug candidates.
• Further experimental validation is recommended for the identified compounds.