How ribesstefano/protac-splitter works
PROTAC-Splitter accepts PROTAC SMILES and returns E3-ligand, linker, and POI-warhead predictions. Its central API supports heuristic, graph/XGBoost, Transformer, and adaptive QC-gated strategies; CLI, HTTP, and Gradio interfaces expose that workflow. Related product capabilities include dataset curation and synthetic-data generation, graph/model training, prediction repair, plausibility evaluation, and visualization. Edges follow the excerpted code, SOURCE INDEX references, or README; unsampled subsystem links are omitted where their caller is not evidenced.
Main entry points: protac_splitter.py Split orchestrator · splitting_algorithms.py Heuristic splitter · edge_classifier.py Graph classifier · curation.py Dataset curation
Components 24
User interfaces
- Python API
protac_splitter/__init__.py - Command line
protac_splitter/cli.py - HTTP service
protac_splitter/serve/app.py - Gradio app
scripts/protac_splitter_app.py
Splitting and inference
- Split orchestrator
protac_splitter/protac_splitter.py - Heuristic splitter
protac_splitter/graphs/splitting_algorithms.py - Graph classifier
protac_splitter/graphs/edge_classifier.py - Graph features
protac_splitter/graphs/edge_features.py - Ligand clustering
protac_splitter/graphs/clustering.py - Transformer inference
protac_splitter/llms/model_utils.py
Chemistry and quality
- Chemoinformatics
protac_splitter/chemoinformatics.py - Prediction repair
protac_splitter/fixing_functions.py - Split evaluation
protac_splitter/evaluation.py - Molecule display
protac_splitter/display_utils.py
Dataset workflows
- Dataset curation
protac_splitter/data/curation/curation.py - Substructure extraction
protac_splitter/data/curation/substructure_extraction.py - Bond adjustments
protac_splitter/data/curation/bond_adjustments.py - Data generation
protac_splitter/data/generation/generation.py - Functional groups
protac_splitter/data/generation/functional_groups.py
Model development
- LLM data prep
protac_splitter/llms/data_utils.py - LLM training
protac_splitter/llms/training.py - LLM evaluation
protac_splitter/llms/evaluation.py
Other components
- SMILES user
- Hugging Face Hub
Connections 27
- SMILES user → Python API: submits SMILES
README.md - SMILES user → Command line: submits SMILES
README.md - SMILES user → HTTP service: sends requests
protac_splitter/serve/app.py - SMILES user → Gradio app: submits SMILES
README.md - Command line → Split orchestrator: calls
protac_splitter/cli.py - HTTP service → Split orchestrator: calls
protac_splitter/serve/app.py - Gradio app → Split orchestrator: calls
scripts/protac_splitter_app.py - Python API → Split orchestrator: exposes
README.md - Split orchestrator → Heuristic splitter: dispatches
protac_splitter/protac_splitter.py - Split orchestrator → Graph classifier: dispatches
protac_splitter/protac_splitter.py - Split orchestrator → Transformer inference: dispatches
protac_splitter/protac_splitter.py - Split orchestrator → Prediction repair: repairs predictions
protac_splitter/protac_splitter.py - Split orchestrator → Split evaluation: scores candidates
protac_splitter/protac_splitter.py - Heuristic splitter → Graph classifier: uses classifier
protac_splitter/graphs/splitting_algorithms.py - Graph classifier → Graph features: builds features
protac_splitter/graphs/edge_classifier.py - Heuristic splitter → Ligand clustering: uses representatives
protac_splitter/graphs/splitting_algorithms.py - Heuristic splitter → Bond adjustments: adjusts fragments
protac_splitter/graphs/splitting_algorithms.py - Split orchestrator → Chemoinformatics: uses utilities
protac_splitter/protac_splitter.py - Prediction repair → Substructure extraction: uses extraction
protac_splitter/fixing_functions.py - Dataset curation → Substructure extraction: extracts substructures
protac_splitter/data/curation/curation.py - Dataset curation → Bond adjustments: adjusts bonds
protac_splitter/data/curation/curation.py - Data generation → Split evaluation: checks outputs
protac_splitter/data/generation/generation.py - Data generation → Functional groups: uses groups
protac_splitter/data/generation/__init__.py - LLM training → LLM data prep: prepares data
protac_splitter/llms/training.py - LLM training → LLM evaluation: evaluates model
protac_splitter/llms/training.py - Transformer inference → Hugging Face Hub: loads pretrained model
README.md - Graph classifier → Hugging Face Hub: downloads classifier
README.md