AI based natural inhibitor targeting RPS20 for colorectal cancer treatment using integrated computational approaches

Area of Science:

• Computational drug discovery
• Oncology
• Bioinformatics

Background:

• Global cancer incidence necessitates novel therapeutic strategies.
• Machine learning and artificial intelligence accelerate the identification of anticancer targets and drug development.
• Natural compounds offer a rich source for novel drug discovery.

Purpose of the Study:

• To explore natural compounds as potential treatments for colorectal cancer.
• To identify and validate novel anticancer targets using computational approaches.
• To discover a potent natural lead drug candidate for colorectal cancer.

Main Methods:

• Sourcing 20 natural compounds from PubChem and obtaining RPS20 protein structure from AlphaFold.
• Protein structure validation using Ramachandran plot and ERRAT, and binding site identification using DeepSite.
• Virtual screening, ADMET analysis, AI-based drug design (WADDAICA), molecular docking, molecular dynamics (MD) simulations, MMGBSA, and DFT calculations.

Main Results:

• Indirubin was identified as the lead drug candidate with favorable ADMET criteria and binding affinity.
• AI-based drug design and subsequent validation confirmed Indirubin's potential.
• DFT calculations revealed a moderate HOMO-LUMO gap, suggesting potential reactivity and biological target interaction.

Conclusions:

• Indirubin shows significant potential as a potent and effective colorectal cancer inhibitor.
• The study highlights the efficacy of computational approaches and AI in accelerating drug discovery.
• Indirubin may offer a therapeutic option with high efficacy and minimal side effects.