How mwasifanwar/geneforge works
Project type and purpose
• This is a Python-based computational pipeline (“geneforge”) for designing novel protein structures.
• It exposes an API, processes input sequences/structures, applies transformer-based neural models, runs evolutionary algorithms, validates with molecular dynamics, predicts drug-discovery properties, and generates visualizations.
File structure
• Top level:
• main.py – orchestrator
• config.yaml – runtime/config parameters
• src/api/server.py – REST or RPC API entrypoint
• requirements.txt – dependencies
• src/data_processing – sequence/structure parsing and encoding
• src/neural_networks – transformer and property/structure predictors
• src/evolutionary – genetic algorithm, fitness functions, mutation operators
• src/molecular_dynamics – simulation driver, force‐field definitions, analysis
• src/drug_discovery – ADMET, binding‐affinity, docking predictors
• src/visualization – sequence and structure visualization
• src/utils – config loader, bioinformatics helper functions
• tests – unit tests for key modules
README insights
• “Transformer-based system that designs novel protein structures for drug discovery and synthetic biology using evolutionary algorithms and deep learning.”