CrypToth: Cryptic Pocket Detection through Mixed-Solvent Molecular Dynamics Simulations-Based Topological Data...

Area of Science:

• Biochemistry
• Computational Biology
• Drug Discovery

Background:

• Proteins can change shape to reveal hidden binding sites, known as cryptic sites.
• These cryptic sites are important for drug discovery but are challenging to predict accurately.

Purpose of the Study:

• To introduce CrypToth, an advanced computational method for precise identification of cryptic sites.
• To improve the prediction accuracy of cryptic sites for drug discovery applications.

Main Methods:

• Integrated topological data analysis (persistent homology) with mixed-solvent molecular dynamics (MSMD) simulations.
• Utilized six different chemical probes in MSMD simulations to identify potential binding hotspots.
• Applied topological data analysis to rank hotspots based on their likelihood of being cryptic sites.

Main Results:

• CrypToth demonstrated superior performance in identifying cryptic sites compared to existing machine-learning methods.
• In evaluations with nine target proteins, CrypToth successfully ranked the cryptic site hotspots highest in seven cases.
• The method effectively explores protein surface hotspots and assesses conformational variability for accurate cryptic site prediction.

Conclusions:

• CrypToth offers a synergistic approach combining MSMD simulations and topological data analysis for high-accuracy cryptic site prediction.
• This method enhances the ability to identify and target cryptic sites, advancing drug discovery efforts.
• CrypToth provides a powerful tool for understanding protein conformational dynamics and ligand interactions.