DeepConf

DeepConf is a tool for exploring low-energy conformations of small molecules using machine learning potentials. It leverages ANI-ML, RDKit, TorchANI, and optionally Gaussian16 to generate bioactive conformers for downstream tasks like docking or MD simulations.

---

📘 Citation

DeepConf: Leveraging ANI-ML Potentials for Exploring Local Minima with Application to Bioactive Conformations
Omer Tayfuroglu, Irem Nur Zengin, Mehmet Serdar Koca, Abdulkadir Kocak
Journal of Chemical Information and Modeling, 2024
DOI: 10.1021/acs.jcim.4c02053

---

✅ Requirements

• Conda with Python ≥ 3.8
• RDKit
• ASE
• TorchANI
• PyTorch ≥ 0.4.1
• (Optional): Gaussian16 for QM-level optimization

💡 Tip: Use a GPU for faster ANI optimization.

---

⚙️ Installation

conda create --name DeepConf python=3.8
conda activate DeepConf
conda install -c conda-forge numpy pandas tqdm rdkit ase pytorch torchani dftd3-python

For G16 support:

Ensure Gaussian16 is installed and g16 command is available in your $PATH

git clone https://github.com/otayfuroglu/DeepConf.git

🚀 How to Use

Setup

mkdir workdir
cd workdir
mkdir structures

Copy your ligands:

cp /path/to/*.sdf structures/

Copy and configure the script:

cp /path/to/DeepConf/runDeepConf.sh .

Edit runDeepConf.sh:

ligPrep_DIR="/path/to/DeepConf"
PYTHON_DIR="/path/to/conda/envs/DeepConf/bin"
struct_dir="structures"

Run

bash runDeepConf.sh

🔧 Parameters (for runConfGen.py)

Parameter	Description
structure_dir	Input directory of ligands
ignore_hydrogen	Add hydrogens? (yes/no)
calculator_type	ani2x, g16, uff
optimization_method	E.g. BFGS, LBFGS
optimization_conf	Optimize conformers (yes/no)
optimization_lig	Global ligand optimization (yes/no)
pre_optimization_lig	Pre-conformer ligand opt (yes/no)
genconformer	Generate conformers (yes/no)
nprocs	Number of CPUs
thr_fmax	Force threshold (default: 0.05)
maxiter	Max optimization steps
ETKDG	Use RDKit ETKDG (yes/no)
num_conformers	Number to generate
max_attempts	Max embedding attempts
prune_rms_thresh	RMSD pruning threshold
opt_prune_rms_thresh	Post-opt RMSD threshold
opt_prune_diffE_thresh	Energy diff threshold (eV)
nfold	Pre-prune conformer multiplier
npick	Final conformer count
nscale	Energy scale factor

📂 Output

Each ligand generates its own folder with outputs:

ligand1/
├── pre_opt_ligand1.sdf
├── pre_opt_ligand1_energy.txt
├── minE_conformer.sdf
├── global_ligand1.sdf
├── global_ligand1_energy.txt

Also:

• timings.csv – Processing time per ligand
• failed_files.csv – Ligands that failed during execution

📌 Summary

• ⚡ ANI-ML (ANI2x) conformer optimization
• 🧪 RDKit + ETKDG conformer generation
• 🧠 Pruning based on RMSD and energy thresholds
• 🧬 UFF and G16 support
• 🔗 Easy integration into pipelines