Artificial Intelligence Approach to Find Lead Compounds for Treating Tumors

Area of Science:

• Computational chemistry and drug discovery
• Molecular modeling and simulation
• Peptide-based therapeutics

Background:

• Matrix metalloproteinase 13 (MMP13) is implicated in numerous cancer types.
• Initial screening of traditional Chinese medicine (TCM) compounds using molecular docking and machine learning showed promise but lacked stability in molecular dynamics simulations.
• Peptide design emerged as a viable alternative strategy for developing novel anti-cancer agents.

Purpose of the Study:

• To identify and design stable peptide inhibitors targeting cancer-associated enzymes.
• To evaluate the binding affinity and stability of designed peptides using computational methods.
• To explore the potential of peptide therapeutics in oncology.

Main Methods:

• Screening of traditional Chinese medicine databases and peptide databases using structure-based and ligand-based drug design.
• Validation of molecular docking results using machine learning models (Random Forest, AdaBoost Regressor, Gradient Boosting Regressor) and Deep Learning.
• Molecular dynamics (MD) simulations (300 ns) to assess the stability of receptor-peptide complexes.
• ZDOCK protocol and Refine Docked protein protocol for affinity assessment.

Main Results:

• Machine learning models achieved high accuracy (R² up to 0.922 training, 0.804 test for RF; 0.90 training, 0.810 test for DL) in predicting drug activity for TCM compounds.
• Designed peptides S2, S3, S5, and S6 demonstrated stable binding with target proteins (BCL2, CDK6, MDM2).
• Peptide S2 exhibited a double-site effect with BCL2; S3 competed with cyclin for CDK6 binding; S5 and S6, derived from P53, showed stable binding with MDM2, potentially enhanced by conformational flexibility.

Conclusions:

• Peptide design offers a promising avenue for developing stable anti-cancer agents, overcoming limitations of small molecules.
• The designed peptides S2, S3, S5, and S6 show significant potential as therapeutic candidates for cancer treatment.
• Further biological validation is required to confirm the efficacy and therapeutic potential of these novel peptide inhibitors.