Alterlab Torchdrug

---
name: alterlab-torchdrug
description: Builds PyTorch-native graph neural networks with TorchDrug for molecules and proteins, exposing custom GNN architectures, task/dataset abstractions, molecular generation, retrosynthesis planning, and knowledge-graph reasoning. Use when a project specifically needs TorchDrug's datasets and tasks — GearNet protein-structure models, center-identification/synthon-completion retrosynthesis, or Hetionet knowledge-graph baselines. TorchDrug has had no release since 0.2.1 (July 2023) and needs Python < 3.11, so for new custom GNN work prefer alterlab-torch-geometric, and for ready-made featurizers, MoleculeNet benchmarks, and pre-trained models prefer alterlab-deepchem. Part of the AlterLab Academic Skills suite.
license: Apache-2.0
allowed-tools: Read Write Edit Bash(python:*) Bash(uv:*)
compatibility: "Self-contained — runs locally, no API key or account required. TorchDrug 0.2.1 (unmaintained, last release July 2023) requires Python >=3.7,<3.11 (use `uv venv --python 3.10`) plus torch-scatter/torch-cluster built for the installed torch and a C++ toolchain for its JIT extensions."
metadata:
    skill-author: AlterLab
    version: "1.1.0"
    last_updated: "2026-09-23"
---

# TorchDrug

## Overview

TorchDrug is a comprehensive PyTorch-based machine learning toolbox for drug discovery and molecular science. Apply graph neural networks, pre-trained models, and task definitions to molecules, proteins, and biological knowledge graphs, including molecular property prediction, protein modeling, knowledge graph reasoning, molecular generation, retrosynthesis planning, with 40+ curated datasets and 20+ model architectures.

## When to Use This Skill

This skill should be used when working with:

**Data Types:**
- SMILES strings or molecular structures
- Protein sequences or 3D structures (PDB files)
- Chemical reactions and retrosynthesis
- Biomedical knowledge graphs
- Drug discovery datasets

**Tasks:**
- Predicting molecular properties (solubility, toxicity, activity)
- Protein function or structure prediction
- Drug-target binding prediction
- Generating new molecular structures
- Planning chemical synthesis routes
- Link prediction in biomedical knowledge bases
- Training graph neural networks on scientific data

**Libraries and Integration:**
- TorchDrug is the primary library
- Often used with RDKit for cheminformatics
- Compatible with PyTorch and PyTorch Lightning
- Integrates with AlphaFold and ESM for proteins

### Does NOT Trigger

| Scenario | Use Instead |
|----------|-------------|
| Building a general custom GNN today (active library, current Python/torch) | `alterlab-torch-geometric` |
| Ready-made featurizers, MoleculeNet benchmarks, and pretrained models with less code | `alterlab-deepchem` |
| Only turning molecules into feature matrices for your own models | `alterlab-molfeat` |
| Only sourcing a labeled benchmark dataset (ADMET/DTI) with official splits | `alterlab-pytdc` |
| Querying an existing biomedical knowledge graph rather than training a KG model | `alterlab-primekg` |

## Getting Started

### Installation

```bash
# TorchDrug 0.2.1 (last release, Jul 2023) requires Python >=3.7,<3.11 and
# torch >=1.8. It will NOT solve on Python 3.11+ — pin an older interpreter:
uv venv --python 3.10
uv pip install torch                        # install torch first (see below)
# torch-scatter / torch-cluster ship only as sdists, so they compile against the
# installed torch — either install from the PyG wheel index or ensure a C++ toolchain:
uv pip install torch-scatter torch-cluster \
  --find-links "https://data.pyg.org/whl/torch-${TORCH_VERSION}+${CUDA}.html"  # e.g. cpu, cu121
uv pip install torchdrug==0.2.1
```

Gotchas:
- Install `torch` first; TorchDrug and `torch-scatter`/`torch-cluster` build graph ops against the installed torch, and mismatched builds fail at import.
- TorchDrug vendors its own `data.DataLoader`, `data.Graph`, and `core.Engine` — reach for those, not the bare PyTorch equivalents (see the loop below).
- TorchDrug depends on `rdkit-pypi`, which has no wheels past 2022.09 (Python ≤ 3.11) — another reason the stack is pinned to an older interpreter.
- It is unmaintained (no release since 0.2.1, July 2023); for new Python/torch stacks prefer `alterlab-torch-geometric` or `alterlab-deepchem`. Use this skill when you specifically need TorchDrug's task/dataset abstractions (e.g. GearNet, its retrosynthesis pipeline).

### Quick Example

```python
import torch
from torchdrug import data, datasets, models, tasks

# Load molecular dataset. MoleculeNet-style datasets such as BBBP have no .split()
# method (only datasets with predefined splits do) — split explicitly:
dataset = datasets.BBBP("~/molecule-datasets/")
lengths = [int(0.8 * len(dataset)), int(0.1 * len(dataset))]
lengths += [len(dataset) - sum(lengths)]
train_set, valid_set, test_set = data.ordered_scaffold_split(dataset, lengths)
# (or torch.utils.data.random_split(dataset, lengths) for a random split)

# Define GNN model
model = models.GIN(
    input_dim=dataset.node_feature_dim,
    hidden_dims=[256, 256, 256],
    edge_input_dim=dataset.edge_feature_dim,
    batch_norm=True,
    readout="mean"
)

# Create property prediction task
task = tasks.PropertyPrediction(
    model,
    task=dataset.tasks,
    criterion="bce",
    metric=["auroc", "auprc"]
)

# Train with a native PyTorch loop.
# preprocess() builds the task's MLP prediction head (and label statistics) — core.Engine
# calls it for you; in a manual loop call it BEFORE creating the optimizer, or the head
# is missing and its parameters are never optimized.
task.preprocess(train_set, valid_set, test_set)
# NOTE: use torchdrug.data.DataLoader (NOT torch.utils.data.DataLoader) — its
# default graph_collate packs data.Graph objects; the stock PyTorch collate cannot.
optimizer = torch.optim.Adam(task.parameters(), lr=1e-3)
train_loader = data.DataLoader(train_set, batch_size=32, shuffle=True)

for epoch in range(100):
    for batch in train_loader:
        # task.forward returns (loss, metric), not a bare tensor.
        loss, metric = task(batch)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
```

For the standard (non-custom) path, `core.Engine(task, train_set, valid_set, test_set, optimizer, batch_size=...)` wraps this loop and handles batching/devices. Use the manual loop above when you need full control over the training step.

## Core Capabilities

### 1. Molecular Property Prediction

Predict chemical, physical, and biological properties of molecules from structure.

**Use Cases:**
- Drug-likeness and ADMET properties
- Toxicity screening
- Quantum chemistry properties
- Binding affinity prediction

**Key Components:**
- 20+ molecular datasets (BBBP, HIV, Tox21, QM9, etc.)
- GNN models (GIN, GAT, SchNet)
- PropertyPrediction and MultipleBinaryClassification tasks

**Reference:** See `references/molecular_property_prediction.md` for:
- Complete dataset catalog
- Model selection guide
- Training workflows and best practices
- Feature engineering details

### 2. Protein Modeling

Work with protein sequences, structures, and properties.

**Use Cases:**
- Enzyme function prediction
- Protein stability and solubility
- Subcellular localization
- Protein-protein interactions
- Structure prediction

**Key Components:**
- 15+ protein datasets (EnzymeCommission, GeneOntology, PDBBind, etc.)
- Sequence models (ESM, ProteinBERT, ProteinLSTM)
- Structure models (GearNet, SchNet)
- Multiple task types for different prediction levels

**Reference:** See `references/protein_modeling.md` for:
- Protein-specific datasets
- Sequence vs structure models
- Pre-training strategies
- Integration with AlphaFold and ESM

### 3. Knowledge Graph Reasoning

Predict missing links and relationships in biological knowledge graphs.

**Use Cases:**
- Drug repurposing
- Disease mechanism discovery
- Gene-disease associations
- Multi-hop biomedical reasoning

**Key Components:**
- General KGs (FB15k, WN18) and biomedical (Hetionet)
- Embedding models (TransE, RotatE, ComplEx)
- KnowledgeGraphCompletion task

**Reference:** See `references/knowledge_graphs.md` for:
- Knowledge graph datasets (including Hetionet with 45k biomedical entities)
- Embedding model comparison
- Evaluation metrics and protocols
- Biomedical applications

### 4. Molecular Generation

Generate novel molecular structures with desired properties.

**Use Cases:**
- De novo drug design
- Lead optimization
- Chemical space exploration
- Property-guided generation

**Key Components:**
- Autoregressive generation
- GCPN (policy-based generation)
- GraphAutoregressiveFlow
- Property optimization workflows

**Reference:** See `references/molecular_generation.md` for:
- Generation strategies (unconditional, conditional, scaffold-based)
- Multi-objective optimization
- Validation and filtering
- Integration with property prediction

### 5. Retrosynthesis

Predict synthetic routes from target molecules to starting materials.

**Use Cases:**
- Synthesis planning
- Route optimization
- Synthetic accessibility assessment
- Multi-step planning

**Key Components:**
- USPTO-50k reaction dataset
- CenterIdentification (reaction center prediction)
- SynthonCompletion (reactant prediction)
- End-to-end Retrosynthesis pipeline

**Reference:** See `references/retrosynthesis.md` for:
- Task decomposition (center ID → synthon completion)
- Multi-step synthesis planning
- Commercial availability checking
- Integration with other retrosynthesis tools

### 6. Graph Neural Network Models

Comprehensive catalog of GNN architectures for different data types and tasks.

**Available Models:**
- General GNNs: GCN, GAT, GIN, RGCN, MPNN
- 3D-aware: SchNet, GearNet
- Protein-specific: ESM, ProteinBERT, GearNet
- Knowledge graph: TransE, RotatE, ComplEx, SimplE
- Generative: GraphAutoregressiveFlow

**Reference:** See `references/models_architectures.md` for:
- Detailed model descriptions
- Model selection guide by task and dataset
- Architecture comparisons
- Implementation tips

### 7. Datasets

40+ curated datasets spanning chemistry, biology, and knowledge graphs.

**Categories:**
- Molecular properties (drug discovery, quantum chemistry)
- Protein properties (function, structure, interactions)
- Knowledge graphs (general and biomedical)
- Retrosynthesis reactions

**Reference:** See `references/datasets.md` for:
- Complete dataset catalog with sizes and tasks
- Dataset selection guide
- Loading and preprocessing
- Splitting strategies (random, scaffold)

## Common Workflows

### Workflow 1: Molecular Property Prediction

**Scenario:** Predict blood-brain barrier penetration for drug candidates.

**Steps:**
1. Load dataset: `datasets.BBBP()`
2. Choose model: GIN for molecular graphs
3. Define task: `PropertyPrediction` with binary classification
4. Train with scaffold split for realistic evaluation
5. Evaluate using AUROC and AUPRC

**Navigation:** `references/molecular_property_prediction.md` → Dataset selection → Model selection → Training

### Workflow 2: Protein Function Prediction

**Scenario:** Predict enzyme function from sequence.

**Steps:**
1. Load dataset: `datasets.EnzymeCommission()`
2. Choose model: ESM (pre-trained) or GearNet (with structure)
3. Define task: `PropertyPrediction` with multi-class classification
4. Fine-tune pre-trained model or train from scratch
5. Evaluate using accuracy and per-class metrics

**Navigation:** `references/protein_modeling.md` → Model selection (sequence vs structure) → Pre-training strategies

### Workflow 3: Drug Repurposing via Knowledge Graphs

**Scenario:** Find new disease treatments in Hetionet.

**Steps:**
1. Load dataset: `datasets.Hetionet()`
2. Choose model: RotatE or ComplEx
3. Define task: `KnowledgeGraphCompletion`
4. Train with negative sampling
5. Query for "Compound-treats-Disease" predictions
6. Filter by plausibility and mechanism

**Navigation:** `references/knowledge_graphs.md` → Hetionet dataset → Model selection → Biomedical applications

### Workflow 4: De Novo Molecule Generation

**Scenario:** Generate drug-like molecules optimized for target binding.

**Steps:**
1. Train property predictor on activity data
2. Choose generation approach: GCPN for RL-based optimization
3. Define reward function combining affinity, drug-likeness, synthesizability
4. Generate candidates with property constraints
5. Validate chemistry and filter by drug-likeness
6. Rank by multi-objective scoring

**Navigation:** `references/molecular_generation.md` → Conditional generation → Multi-objective optimization

### Workflow 5: Retrosynthesis Planning

**Scenario:** Plan synthesis route for target molecule.

**Steps:**
1. Load dataset: `datasets.USPTO50k()`
2. Train center identification model (RGCN)
3. Train synthon completion model (GIN)
4. Combine into end-to-end retrosynthesis pipeline
5. Apply recursively for multi-step planning
6. Check commercial availability of building blocks

**Navigation:** `references/retrosynthesis.md` → Task types → Multi-step planning

## Integration Patterns

### With RDKit

Convert between TorchDrug molecules and RDKit:
```python
from torchdrug import data
from rdkit import Chem

# SMILES → TorchDrug molecule
smiles = "CCO"
mol = data.Molecule.from_smiles(smiles)

# TorchDrug → RDKit
rdkit_mol = mol.to_molecule()

# RDKit → TorchDrug
rdkit_mol = Chem.MolFromSmiles(smiles)
mol = data.Molecule.from_molecule(rdkit_mol)
```

### With AlphaFold/ESM

Use predicted structures:
```python
from torchdrug import data, layers
from torchdrug.layers import geometry

# Load AlphaFold predicted structure
protein = data.Protein.from_pdb("AF-P12345-F1-model_v4.pdb")

# Build a residue-level graph with sequential + spatial edges
graph_construction_model = layers.GraphConstruction(
    node_layers=[geometry.AlphaCarbonNode()],
    edge_layers=[
        geometry.SpatialEdge(radius=10.0, min_distance=5),
        geometry.SequentialEdge(max_distance=2),
    ],
    edge_feature="gearnet",
)
# GraphConstruction operates on a packed protein batch, not a bare Protein.
graph = graph_construction_model(data.Protein.pack([protein]))
```

### With PyTorch Lightning

Wrap tasks for Lightning training:
```python
import torch
import pytorch_lightning as pl   # or: import lightning.pytorch as pl

class LightningTask(pl.LightningModule):
    def __init__(self, torchdrug_task):
        super().__init__()
        self.task = torchdrug_task      # call task.preprocess(...) before training

    def training_step(self, batch, batch_idx):
        loss, metric = self.task(batch)  # forward returns (loss, metric)
        return loss

    def validation_step(self, batch, batch_idx):
        pred = self.task.predict(batch)
        target = self.task.target(batch)
        return {"pred": pred, "target": target}

    def configure_optimizers(self):
        return torch.optim.Adam(self.parameters(), lr=1e-3)
```

## Technical Details

For deep dives into TorchDrug's architecture:

**Core Concepts:** See `references/core_concepts.md` for:
- Architecture philosophy (modular, configurable)
- Data structures (Graph, Molecule, Protein, PackedGraph)
- Model interface and forward function signature
- Task interface (predict, target, forward, evaluate)
- Training workflows and best practices
- Loss functions and metrics
- Common pitfalls and debugging

## Quick Reference Cheat Sheet

**Choose Dataset:**
- Molecular property → `references/datasets.md` → Molecular section
- Protein task → `references/datasets.md` → Protein section
- Knowledge graph → `references/datasets.md` → Knowledge graph section

**Choose Model:**
- Molecules → `references/models_architectures.md` → GNN section → GIN/GAT/SchNet
- Proteins (sequence) → `references/models_architectures.md` → Protein section → ESM
- Proteins (structure) → `references/models_architectures.md` → Protein section → GearNet
- Knowledge graph → `references/models_architectures.md` → KG section → RotatE/ComplEx

**Common Tasks:**
- Property prediction → `references/molecular_property_prediction.md` or `references/protein_modeling.md`
- Generation → `references/molecular_generation.md`
- Retrosynthesis → `references/retrosynthesis.md`
- KG reasoning → `references/knowledge_graphs.md`

**Understand Architecture:**
- Data structures → `references/core_concepts.md` → Data Structures
- Model design → `references/core_concepts.md` → Model Interface
- Task design → `references/core_concepts.md` → Task Interface

## Troubleshooting Common Issues

**Issue: Dimension mismatch errors**
→ Check `model.input_dim` matches `dataset.node_feature_dim`
→ See `references/core_concepts.md` → Essential Attributes

**Issue: Poor performance on molecular tasks**
→ Use scaffold splitting, not random
→ Try GIN instead of GCN
→ See `references/molecular_property_prediction.md` → Best Practices

**Issue: Protein model not learning**
→ Use pre-trained ESM for sequence tasks
→ Check edge construction for structure models
→ See `references/protein_modeling.md` → Training Workflows

**Issue: Memory errors with large graphs**
→ Reduce batch size
→ Use gradient accumulation
→ See `references/core_concepts.md` → Memory Efficiency

**Issue: Generated molecules are invalid**
→ Add validity constraints
→ Post-process with RDKit validation
→ See `references/molecular_generation.md` → Validation and Filtering

## Resources

**Official Documentation:** https://torchdrug.ai/docs/
**GitHub:** https://github.com/DeepGraphLearning/torchdrug
**Paper:** TorchDrug: A Powerful and Flexible Machine Learning Platform for Drug Discovery

## Summary

Navigate to the appropriate reference file based on your task:

1. **Molecular property prediction** → `molecular_property_prediction.md`
2. **Protein modeling** → `protein_modeling.md`
3. **Knowledge graphs** → `knowledge_graphs.md`
4. **Molecular generation** → `molecular_generation.md`
5. **Retrosynthesis** → `retrosynthesis.md`
6. **Model selection** → `models_architectures.md`
7. **Dataset selection** → `datasets.md`
8. **Technical details** → `core_concepts.md`

Each reference provides comprehensive coverage of its domain with examples, best practices, and common use cases.