MolCRAFT Series for Drug Design

Welcome to the official repository for the MolCRAFT series of projects! This series focuses on developing and improving deep learning models for structure-based drug design (SBDD) and molecule optimization (SBMO). Our goal is to create molecules with high binding affinity and plausible 3D conformations.

This repository contains the source code for the following projects:

MolCRAFT: Structure-Based Drug Design in Continuous Parameter Space (ICML'24)
MolJO: Empower Structure-Based Molecule Optimization with Gradient Guided Bayesian Flow Networks (ICML'25)
MolPilot: Piloting Structure-Based Drug Design via Modality-Specific Optimal Schedule (ICML'25)

📜 Overview

The MolCRAFT series addresses critical challenges in generative models for SBDD, including modeling molecular geometries, handling hybrid continuous-discrete spaces, and optimizing molecules against protein targets. Each project introduces novel methodologies and achieves state-of-the-art performance on relevant benchmarks.

🧭 Navigation

Folder	TL, DR	Description
MolCRAFT	Unified Space for Molecule Generation	MolCRAFT is the first SBDD generative model based on Bayesian Flow Network (BFN) operating in the unified continuous parameter space for different modalities, with variance reduction sampling strategy to generate high-quality samples with more than 10x speedup.
MolJO	Gradient-Guided Molecule Optimization	MolJO is a gradient-based Structure-Based Molecule Optimization (SBMO) framework derived within BFN. It employs joint guidance across continuous coordinates and discrete atom types, alongside a backward correction strategy for effective optimization.
MolPilot	Optimal Scheduling	MolPilot enhances SBDD by introducing a VLB-Optimal Scheduling (VOS) strategy for the twisted multimodal probability paths, significantly improving molecular geometries and interaction modeling, achieving 95.9% PB-Valid rate.

🚀 Projects

MolCRAFT (Let's Craft the Molecules)

Description: MolCRAFT is the first SBDD model that employs BFN and operates in a continuous parameter space. It introduces a novel noise-reduced sampling strategy to generate molecules with superior binding affinity and more stable 3D structures. MolCRAFT has demonstrated its ability to accurately model interatomic interactions, achieving reference-level Vina Scores.
Key Contributions:
- Operates in continuous parameter space for SBDD within BFN framework.
- Novel variance reduction sampling strategy that improves both sample quality and efficiency.
- Achieves state-of-the-art binding affinity and structural stability.

MolJO (Molecule Joint Optimization)

Description: MolJO is a gradient-based SBMO framework that leverages a continuous and differentiable space derived through Bayesian inference. It facilitates joint guidance signals across different modalities (continuous coordinates and discrete atom types) while preserving SE(3)-equivariance. MolJO introduces a novel backward correction strategy for an effective trade-off between exploration and exploitation.
Key Contributions:
- Gradient-based SBMO framework with joint guidance across different modalities.
- Backward correction strategy for optimized exploration-exploitation.
- State-of-the-art performance in practical optimization tasks, including multi-objective and constrained optimization for R-group redesign, scaffold hopping, etc.

MolPilot (How to Pilot the Aircraft)

Description: MolPilot addresses challenges in geometric structure modeling by focusing on the twisted probability path of multi-modalities (continuous 3D positions and discrete 2D topologies). It proposes a VLB-Optimal Scheduling (VOS) strategy, optimizing the Variational Lower Bound as a path integral for SBDD. MolPilot significantly enhances molecular geometries and interaction modeling.
Key Contributions:
- Addresses multi-modality challenges in SBDD.
- Introduces VLB-Optimal Scheduling (VOS) strategy, generally applicable to a wide range of frameworks including diffusions.
- Achieves 95.9% PoseBusters passing rate on CrossDock with significantly improved molecular geometries.

MolCRAFT

Official implementation of ICML 2024 "MolCRAFT: Structure-Based Drug Design in Continuous Parameter Space".

🎉 Our demo is now available here. Welcome to have a try!

Environment

It is highly recommended to install via docker if a Linux server with NVIDIA GPU is available.

Otherwise, you might check README for env for further details of docker or conda setup.

Prerequisite

A docker with nvidia-container-runtime enabled on your Linux system is required.

This repo provides an easy-to-use script to install docker and nvidia-container-runtime, in ./docker run sudo ./setup_docker_for_host.sh to set up your host machine.

For details, please refer to the install guide.

Install via Docker

We highly recommend you to set up the environment via docker, since all you need to do is a simple make command.

cd ./docker
make

Data

We use the same data as TargetDiff. Data used for training / evaluating the model should be put in the data folder by default, and accessible in the data Google Drive folder.

To train the model from scratch, download the lmdb file and split file into data folder:

crossdocked_v1.1_rmsd1.0_pocket10_processed_final.lmdb
crossdocked_pocket10_pose_split.pt

To evaluate the model on the test set, download and unzip the test_set.zip into data folder. It includes the original PDB files that will be used in Vina Docking.

Training

Run make -f scripts.mk (without the need for data preparation), or alternatively (with data folder correctly configured),

python train_bfn.py --exp_name ${EXP_NAME} --revision ${REVISION}

where the default values should be set the same as:

python train_bfn.py --sigma1_coord 0.03 --beta1 1.5 --lr 5e-4 --time_emb_dim 1 --epochs 15 --max_grad_norm Q --destination_prediction True --use_discrete_t True --num_samples 10 --sampling_strategy end_back_pmf

Testing

For quick evaluation of the official checkpoint, refer to make evaluate in scripts.mk:

python train_bfn.py --test_only --no_wandb --ckpt_path ./checkpoints/${CKPT_NAME}

Debugging

For quick debugging training process, run make debug -f scripts.mk:

python train_bfn.py --no_wandb --debug --epochs 1

Sampling

We provide the pretrained MolCRAFT checkpoint here.

Sampling for pockets in the testset

Run make evaluate -f scripts.mk, or alternatively,

python train_bfn.py --config_file configs/default.yaml --exp_name ${EXP_NAME} --revision ${REVISION} --test_only --num_samples ${NUM_MOLS_PER_POCKET} --sample_steps 100

The output molecules vina_docked.pt for all 100 test pockets will be saved in ./logs/${USER}_bfn_sbdd/${EXP_NAME}/${REVISION}/test_outputs/${TIMESTAMP} folders.

Sampling from pdb file

To sample from a whole protein pdb file, we need the corresponding reference ligand to clip the protein pocket (a 10A region around the reference position).

Below is an example that stores the generated 10 molecules under output folder. The configurations are managed in the call() function of sample_for_pocket.py.

python sample_for_pocket.py ${PDB_PATH} ${SDF_PATH}

Evaluation

Evaluating molecules

For binding affinity (Vina Score / Min / Dock) and molecular properties (QED, SA), it is calculated upon sampling.

For PoseCheck (strain energy, clashes) and other conformational results (bond length, bond angle, torsion angle, RMSD), please refer to test folder.

Evaluating meta files

We provide samples for all SBDD baselines in the sample Google Drive folder.

You may download the all_samples.tar.gz and then tar xzvf all_samples.tar.gz, which extracts all the pt files into samples folder for evaluation.

Citation

@article{qu2024molcraft,
  title={MolCRAFT: Structure-Based Drug Design in Continuous Parameter Space},
  author={Qu, Yanru and Qiu, Keyue and Song, Yuxuan and Gong, Jingjing and Han, Jiawei and Zheng, Mingyue and Zhou, Hao and Ma, Wei-Ying},
  journal={ICML 2024},
  year={2024}
}

@article{song2024unified,
  title={Unified Generative Modeling of 3D Molecules via Bayesian Flow Networks},
  author={Song, Yuxuan and Gong, Jingjing and Qu, Yanru and Zhou, Hao and Zheng, Mingyue and Liu, Jingjing and Ma, Wei-Ying},
  journal={ICLR 2024},
  year={2024}
}