How drewlindsey/denovoprediction works
Main Components Identification
a. Core Pipeline Components (Primary Processing Flow):
• Conformation Initializer
• Fragment Library Generator
• Seef (Energy Computation)
• Conformation Sampler
• Residue Mapper
• PDB Mapper
b. External Dependencies:
• calRW (RW Potential calculation)
• CRANKITE (PDB file format mapping)
• ROBETTA (Fragment generation)
c. Web Interface:
• Flask Web Server
• Redis (indicated by run-redis.sh)
• Celery (indicated by run-celery.sh)
[Flask Web Server] ←→ [Celery Worker] ←→ [Redis]
[Pipeline Controller]
[Conformation Initializer] → [Fragment Library Generator] → [Seef] → [Conformation Sampler] → [Residue Mapper] → [PDB Mapper]
Specific Implementation Details
a. Data Flow:
• How protein sequence data enters the system
• Intermediate data transformations
• Final PDB structure output
b. Module Organization (based on src/ directory):
• Conformation module (src/conformation/)
• Fragments module (src/fragments/)
• Pipeline module (src/pipeline/)
• Score module (src/score/)
• Seef module (src/seef/)
Additional Elements
• Input/Output formats at each stage
• Key file formats (PDB, fragment libraries)
• Processing stages with their respective computational methods
• Integration points with external tools
Notes for Implementation
• Use standard UML or similar notation for components
• Brief descriptions of each component's function
• Error handling and alternative paths
• Parallel processing capabilities (especially for Celery workers)
• Where configuration files are used