How uttkarsh779/knf-molecular-descriptors works
KNF Studio is a desktop application that accepts molecular structure files and presents computed KNF descriptors and interaction visualizations. The main workflow runs from the user’s run-management interface through Electron and the Python API, then through geometry optimization, NCI analysis, descriptor and fingerprint generation, and persisted results to the Results and Explorer views. The README documents the scientific stages; source excerpts confirm selected internal calls. Some UI-to-API and backend orchestration details are not shown, so those links are marked unverified.
Main entry points: knf_core/pipeline.py KNF pipeline · nci_torch/pipeline.py NCI pipeline · RunManager.tsx Run manager · server.py Python API
Components 22
Desktop interface
- Run manager
frontend/src/pages/RunManager.tsx - Electron shell
frontend/electron/main.ts - Backend launcher
frontend/electron/backend.ts - Results table
frontend/src/pages/Results.tsx - Molecule explorer
frontend/src/pages/Explorer.tsx - Analysis view
frontend/src/pages/Analysis.tsx
Run service
- Python API
backend/NCIForge/server.py
Scientific computation
- KNF pipeline
backend/NCIForge/knf_core/pipeline.py - Geometry handling
backend/NCIForge/knf_core/geometry.py - xTB optimizer
backend/NCIForge/knf_core/xtb.py - Structure conversion
backend/NCIForge/knf_core/converter.py - KNF descriptors
backend/NCIForge/knf_core/knf_vector.py - Tool configuration
backend/NCIForge/knf_core/autoconfig.py - KUID fingerprint
backend/NCIForge/knf_core/kuid.py - Quadrant classification
backend/NCIForge/knf_core/main.py
NCI analysis
- NCI pipeline
backend/NCIForge/knf_core/nci_torch/pipeline.py - Molden parser
backend/NCIForge/knf_core/nci_torch/molden.py - Density grid
backend/NCIForge/knf_core/nci_torch/grid.py - NCI engine
backend/NCIForge/knf_core/nci_torch/engine.py - Device router
backend/NCIForge/knf_core/nci_torch/router.py
Results and storage
- Run database
backend/NCIForge/database.py
Other components
- KNF Studio user
Connections 22
- KNF Studio user → Run manager: starts runs
README.md - Run manager → Electron shell: submits run
- Electron shell → Backend launcher: starts backend
frontend/electron/main.ts - Run manager → Python API: requests runs
- Python API → KNF pipeline: dispatches computation
- KNF pipeline → Geometry handling: prepares geometry
backend/NCIForge/knf_core/pipeline.py - KNF pipeline → xTB optimizer: optimizes geometry
backend/NCIForge/knf_core/pipeline.py - KNF pipeline → Structure conversion: converts structures
backend/NCIForge/knf_core/pipeline.py - KNF pipeline → NCI pipeline: runs NCI
backend/NCIForge/knf_core/pipeline.py - NCI pipeline → Molden parser: reads wavefunction
backend/NCIForge/knf_core/nci_torch/pipeline.py - NCI pipeline → Density grid: builds grid
backend/NCIForge/knf_core/nci_torch/pipeline.py - NCI pipeline → NCI engine: computes fields
backend/NCIForge/knf_core/nci_torch/pipeline.py - KNF pipeline → KNF descriptors: computes descriptors
backend/NCIForge/knf_core/pipeline.py - KNF pipeline → Device router: routes device
- KNF pipeline → Tool configuration: configures tools
- KNF pipeline → KUID fingerprint: creates fingerprint
- KNF pipeline → Quadrant classification: classifies results
backend/NCIForge/knf_core/main.py - Python API → Run database: reads and writes
backend/NCIForge/server.py - Run database → Results table: serves run results
- Results table → KNF Studio user: shows descriptors
README.md - Molecule explorer → KNF Studio user: plots molecules
README.md - Analysis view → KNF Studio user: presents analysis