Computational protein/drug design

Generative ML applied to protein engineering - designing new sequences and validating them computationally before any wet-lab work. Each post follows a real pipeline: sequence/structure generation → filtering → structure prediction → confidence scoring. So far: redesigning a thermostable enzyme with ProteinMPNN inverse folding (validated with AlphaFold2), and screening 24,000 antimicrobial peptide sequences down to 47 high-confidence drug candidates using ML classification + ColabFold.

FarhansBioAI

•2 posts•Since Jul 4, 2026