Title will be here

Filipp Nikitin^1,2   ·   Dylan M. Anstine^2,3   ·   Olexandr Isayev^1,2,4*
1Ray and Stephanie Lane Computational Biology Department, Carnegie Mellon University, Pittsburgh, PA, USA
²Department of Chemistry, Carnegie Mellon University, Pittsburgh, PA, USA
³Department of Chemical Engineering and Materials Science, Michigan State University, East Lansing, MI, USA
⁴Department of Materials Science and Engineering, Carnegie Mellon University, Pittsburgh, PA, USA

📄 Paper   ·   📖 Citation   ·   ⚙️ Setup   ·   🔗 GitHub

^*Corresponding author: olexandr@olexandrisayev.com

---

Overview

Abstract

---

Key Features

---

Setup

This setup has been tested on Ubuntu 22.04, but can be used across multiple platforms as PyTorch, Pytorch Geometric, and RdKit are widely supported. Installation will usually take up to 20 minutes.

Prerequisites

• Python 3.10+
• CUDA-compatible GPU (recommended for training)
• Conda or Mamba (recommended)

Environment Setup

# Clone the repository
git clone https://github.com/isayevlab/LoQI.git
cd LoQI

# Create and activate conda environment
conda create -n loqi python=3.10 -y
conda activate loqi

# Install dependencies
pip install -r requirements.txt
pip install -e .

Data Setup

The training and evaluation require the ChEMBL3D dataset.

Available with this release:

---

Usage

Make sure the package is installed locally: pip install -e .

Model Training

# Train transition state model from scratch
python scripts/train.py \
    --config-path=./conf/ \
    --config-name ts_extended_data \
    train.gpus=1 \
    train.seed=28 \
    run_name=test_train \
    outdir="../test_runs" \
    data.dataset_root="/path/to/ts_dataset"

# Resume training from checkpoint
python scripts/train.py \
    --config-path=./conf/ \
    --config-name ts_extended_data \
    train.gpus=1 \
    train.seed=28 \
    run_name=test_train \
    outdir="../test_runs" \
    resume="./models/last_converted.ckpt"

# Customize training parameters
python scripts/train.py \
    --config-path=./conf/ \
    --config-name ts_extended_data \
    outdir=./outputs \
    train.gpus=2 \
    train.n_epochs=800 \
    train.seed=42 \
    data.batch_size=150 \
    optimizer.lr=0.0001

Model Inference and Sampling

Transition States Generation

# Generate transition states from atom-mapped SMILES
python scripts/sample_transition_state.py \
    --reactant_smi "[C:1]([c:2]1[n:3][o:4][n:5][n:6]1)([H:7])([H:8])[H:9]" \
    --product_smi "[C:1]1([H:7])([H:8])/[C:2](=[N:3]\\[H:9])[N:6]1[N:5]=[O:4]" \
    --config scripts/conf/ts_extended_data.yaml \
    --ckpt models/last_converted.ckpt \
    --output output.xyz \
    --n_samples 1 \
    --batch_size 32

# Generate transition states from XYZ files
python scripts/sample_transition_state.py \
    --reactant_xyz reactant.xyz \
    --product_xyz product.xyz \
    --config scripts/conf/ts_extended_data.yaml \
    --ckpt models/last_converted.ckpt \
    --output output.xyz \
    --n_samples 5 \
    --batch_size 32

# Generate multiple samples per reaction
python scripts/sample_transition_state.py \
    --reactant_smi "[C:1][C:2]([H:3])([H:4])[H:5]" \
    --product_smi "[C:1]=[C:2]([H:3])[H:4]" \
    --config scripts/conf/ts_extended_data.yaml \
    --ckpt models/last_converted.ckpt \
    --output ts_samples.xyz \
    --n_samples 10 \
    --batch_size 32

Input formats:

• SMILES: Atom-mapped SMILES with explicit hydrogens (e.g., [C:1][H:2])
• XYZ: Standard XYZ coordinate files (bonds will be inferred using OpenBabel)

Notes:

• SMILES must have explicit hydrogens and can use atom mapping to specify atom correspondence
• Reactant and product must have the same number of atoms
• Output is saved as XYZ file(s) with transition state coordinates

Available Configurations

Transition State Configs:

• ts_extended_data.yaml - Transition state model configuration
• ts1x.yaml - Alternative transition state configuration

Model Configs:

• loqi.yaml - LoQI stereochemistry-aware conformer generation model
• nextmol.yaml - Alternative configuration for NextMol-style generation

Training Configuration

You can easily override configuration parameters:

# Example with custom parameters
python scripts/train.py \
    --config-path=./conf/ \
    --config-name ts_extended_data \
    outdir=./my_training \
    run_name=my_experiment \
    train.gpus=4 \
    train.n_epochs=500 \
    data.batch_size=64 \
    data.dataset_root="/path/to/ts_dataset" \
    wandb_params.mode=online

---

Citation

# citation is coming sson
@article{
}

You may also find useful our paper and model for low-energy conformer generation:

@article{nikitin2025scalable,
  title={Scalable Low-Energy Molecular Conformer Generation with Quantum Mechanical Accuracy},
  author={Nikitin, Filipp and Anstine, Dylan M and Zubatyuk, Roman and Paliwal, Saee Gopal and Isayev, Olexandr},
  year={2025}
}

This work builds upon the Megalodon architecture. If you use the underlying architecture, please also cite:

@article{reidenbach2025applications,
  title={Applications of Modular Co-Design for De Novo 3D Molecule Generation},
  author={Reidenbach, Danny and Nikitin, Filipp and Isayev, Olexandr and Paliwal, Saee},
  journal={arXiv preprint arXiv:2505.18392},
  year={2025}
}