Combining Molecular Dynamics and Machine Learning to Predict Drug Resistance Causing Variants of BRAF in Colorectal...

Area of Science:

• Oncology
• Molecular Biology
• Computational Chemistry

Background:

• BRAF protein mutations, especially V600E, are prevalent in colorectal cancer (CRC).
• BRAF mutations confer poor prognosis and present therapeutic challenges due to primary and acquired drug resistance.
• Understanding BRAF structural changes is crucial for overcoming treatment resistance.

Purpose of the Study:

• To investigate structural alterations in BRAF mutations linked to drug resistance.
• To identify key residues involved in BRAF inhibitor resistance mechanisms.
• To develop predictive models for BRAF mutation-driven resistance phenotypes.

Main Methods:

• Replica exchange molecular dynamics simulations.
• Machine learning techniques for data analysis.
• Analysis of conformational changes in BRAF proteins.

Main Results:

• Specific conformational changes in mutant BRAF proteins correlate with drug sensitivity/resistance.
• Key residues (e.g., psi494, phi600 for dabrafenib; psi450, phi484 for vemurafenib) were identified.
• These residues are located in the ATP-binding N-lobe of CR3, regulating kinase activity.

Conclusions:

• Structural insights into BRAF mutations explain drug resistance mechanisms.
• Predictive models can guide personalized therapeutic strategies for drug-resistant CRC.
• Targeting specific BRAF variants offers potential for improved CRC treatment outcomes.