torch-molecule 0.1.0

Deep learning packages for molecular discovery with a simple sklearn-style interface

Deep learning for molecular discovery with a simple sklearn-style interface

---

torch-molecule is a package under active development that facilitates molecular discovery through deep learning, featuring a user-friendly, sklearn-style interface. It includes model checkpoints for efficient deployment and benchmarking across a range of molecular tasks. Currently, the package focuses on three main components: Predictive Models, Generative Models, and Representation Models. See the List of Supported Models section for all available models.

Installation

1. Create a Conda environment:

   conda create --name torch_molecule python=3.11.7
   conda activate torch_molecule
   

2. Install using pip (0.1.0):

   pip install torch-molecule
   

3. Install from source for the latest version:

   Clone the repository:

   git clone https://github.com/liugangcode/torch-molecule
   cd torch-molecule
   

   Install in editable mode:

   pip install .
   

Additional Packages

Model	Required Packages
HFPretrainedMolecularEncoder	transformers
BFGNNMolecularPredictor	torch-scatter
GRINMolecularPredictor	torch-scatter

Usage

Refer to the tests folder for more use cases.

Python API Example

The following example demonstrates how to use the GREAMolecularPredictor class from torch_molecule:

More examples could be found in the folders examples and tests.

from torch_molecule import GREAMolecularPredictor

# Train GREA model
grea_model = GREAMolecularPredictor(
    num_task=num_task,
    task_type="regression",
    model_name="GREA_multitask",
    evaluate_criterion='r2',
    evaluate_higher_better=True,
    verbose=True
)

# Fit the model
X_train = ['C1=CC=CC=C1', 'C1=CC=CC=C1']
y_train = [[0.5], [1.5]]
X_val = ['C1=CC=CC=C1', 'C1=CC=CC=C1']
y_val = [[0.5], [1.5]]
N_trial = 10

grea_model.autofit(
    X_train=X_train.tolist(),
    y_train=y_train,
    X_val=X_val.tolist(),
    y_val=y_val,
    n_trials=N_trial,
)

Checkpoints

torch-molecule provides checkpoint functions that can be interacted with on Hugging Face.

from torch_molecule import GREAMolecularPredictor
from sklearn.metrics import mean_absolute_error

# Define the repository ID for Hugging Face
repo_id = "user/repo_id"

# Initialize the GREAMolecularPredictor model
model = GREAMolecularPredictor()

# Train the model using autofit
model.autofit(
    X_train=X.tolist(),  # List of SMILES strings for training
    y_train=y_train,     # numpy array [n_samples, n_tasks] for training labels
    X_val=X_val.tolist(),# List of SMILES strings for validation
    y_val=y_val,         # numpy array [n_samples, n_tasks] for validation labels
)

# Make predictions on the test set
output = model.predict(X_test.tolist()) # (n_sample, n_task)

# Calculate the mean absolute error
mae = mean_absolute_error(y_test, output['prediction'])
metrics = {'MAE': mae}

# Save the trained model to Hugging Face
model.save_to_hf(
    repo_id=repo_id,
    task_id=f"{task_name}",
    metrics=metrics,
    commit_message=f"Upload GREA_{task_name} model with metrics: {metrics}",
    private=False
)

# Load a pretrained checkpoint from Hugging Face
model = GREAMolecularPredictor()
model.load_from_hf(repo_id=repo_id, local_cache=f"{model_dir}/GREA_{task_name}.pt")

# Set model parameters
model.set_params(verbose=True)

# Make predictions using the loaded model
predictions = model.predict(smiles_list)

List of Supported Models

Predictive Models

Model	Reference
GRIN	Learning Repetition-Invariant Representations for Polymer Informatics. May 2025
BFGNN	Graph neural networks extrapolate out-of-distribution for shortest paths. March 2025
SGIR	Semi-Supervised Graph Imbalanced Regression. KDD 2023
GREA	Graph Rationalization with Environment-based Augmentations. KDD 2022
DIR	Discovering Invariant Rationales for Graph Neural Networks. ICLR 2022
SSR	SizeShiftReg: a Regularization Method for Improving Size-Generalization in Graph Neural Networks. NeurIPS 2022
IRM	Invariant Risk Minimization (2019)
RPGNN	Relational Pooling for Graph Representations. ICML 2019
GNNs	Graph Convolutional Networks. ICLR 2017 and Graph Isomorphism Network. ICLR 2019
Transformer (SMILES)	Transformer (Attention is All You Need. NeurIPS 2017) based on SMILES strings
LSTM (SMILES)	Long short-term memory (Neural Computation 1997) based on SMILES strings

Generative Models

Model	Reference
Graph DiT	Graph Diffusion Transformers for Multi-Conditional Molecular Generation. NeurIPS 2024
DiGress	DiGress: Discrete Denoising Diffusion for Graph Generation. ICLR 2023
GDSS	Score-based Generative Modeling of Graphs via the System of Stochastic Differential Equations. ICML 2022
MolGPT	MolGPT: Molecular Generation Using a Transformer-Decoder Model. Journal of Chemical Information and Modeling 2021
JTVAE	Junction Tree Variational Autoencoder for Molecular Graph Generation. ICML 2018.
GraphGA	A Graph-Based Genetic Algorithm and Its Application to the Multiobjective Evolution of Median Molecules. Journal of Chemical Information and Computer Sciences 2004
LSTM (SMILES)	Long short-term memory (Neural Computation 1997) based on SMILES strings

Representation Models

Model	Reference
MoAMa	Motif-aware Attribute Masking for Molecular Graph Pre-training. LoG 2024
GraphMAE	GraphMAE: Self-Supervised Masked Graph Autoencoders. KDD 2022
AttrMasking	Strategies for Pre-training Graph Neural Networks. ICLR 2020
ContextPred	Strategies for Pre-training Graph Neural Networks. ICLR 2020
EdgePred	Strategies for Pre-training Graph Neural Networks. ICLR 2020
InfoGraph	InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization. ICLR 2020
Supervised	Supervised pretraining
Pretrained	More than ten pretrained models from Hugging Face

Project Structure

See the structure of torch_molecule with the command tree -L 2 torch_molecule -I '__pycache__|*.pyc|*.pyo|.git|old*'

Plan

1. Predictive Models: Done: GREA, SGIR, IRM, GIN/GCN w/ virtual, DIR. SMILES-based LSTM/Transformers. TODO more
2. Generative Models: Done: Graph DiT, GraphGA, DiGress, GDS, MolGPT TODO: more
3. Representation Models: Done: MoAMa, AttrMasking, ContextPred, EdgePred. Many pretrained models from HF. TODO: checkpoints, more

Note: This project is in active development, and features may change.

Acknowledgements

The project template was adapted from https://github.com/lwaekfjlk/python-project-template. We thank the authors for their contribution to the open-source community.