Use OpenConf Online
Commercially Available OpenConf No-Code Web Server
OpenConf: Monte Carlo Conformer Generation for Diverse Small-Molecule Ensembles
OpenConf is an open-source, Monte Carlo-based conformer generation package that builds on RDKit to produce diverse conformer ensembles for drug-like molecules. It was created at Rowan, whose newsletter post by Corin Wagen, Nick Casetti, Jonathon Vandezande and Eli Mann introduces it. OpenConf is written in modern Python with few dependencies and installs with uv pip install openconf. It is now the default in Rowan's conformer-search and solvent-dependent-conformer workflows.
Many computed molecular properties depend on finding the low-energy conformers of a flexible molecule. The post reports that RDKit's ETKDG can miss conformers or return high-energy ones for large systems, often favoring twist boats over chairs, and that thousands of conformers may be needed with many converging on the same few. CREST's iMTD-GC works well with careful settings but struggled with quick settings for drug-like molecules in Rowan's high-throughput use, including segfaults and stochastic errors at scale.
Property | Detail |
|---|---|
Approach | Monte Carlo moves on ETKDG seeds, MMFF94 optimization |
Built on | RDKit |
Language | Python, few dependencies |
Metal complexes | UFF with restraints against metal-ligand dissociation |
Macrocycles | Wider energy window, optional low-mode search |
Repository | github.com/rowansci/openconf |
How OpenConf Generates Conformer Ensembles
Seed pool: start from conformers generated by ETKDG.
Monte Carlo moves: add conformers using single and multi-torsion, correlated torsion, global shake, ring flip, crankshaft, kinematic inverse closure and macrocyclic amide flip moves, each followed by MMFF94 optimization.
Adaptive sampling: the success rate of each move is tracked and move probabilities adjust accordingly.
Convergence and deduplication: the pool is deduplicated continuously, and convergence tracking allows early stopping. Batching added to PRISM Pruner makes deduplication about 15x faster.
OpenConf Benchmarks and Performance
The post reports that OpenConf is slower than ETKDG at 10 conformers but noticeably faster at 50 or more, and that it matches or outperforms ETKDG at the same budget on flexible drug-like molecules from the OpenEye Iridium dataset.
Test | Result |
|---|---|
pKa on 16 drug-like molecules | 190 s vs. 846 s with ETKDG, with lower mean error and fewer outliers |
logP on 15 flexible ChEMBL molecules | About 2x faster than previous iMTD-GC settings, virtually identical results |
Macrocycle AMG 410 | 177 unique conformers in 12 seconds |
What is Tamarind Bio?
Tamarind Bio is a no-code bioinformatics platform built to give life scientists and researchers access to powerful computational tools. Many cutting-edge machine learning models are hard to deploy and use. Tamarind provides an intuitive, web-based environment that removes the complexity of high-performance computing, software dependencies and command-line interfaces.
The platform is designed for biologists, chemists and other researchers who may not have a background in programming or cloud infrastructure but want to run models on their own data. Key features include:
A user-friendly graphical interface for setting up and launching experiments
A robust API for integration into existing research pipelines
An automated system for managing and scaling computational resources
Tamarind treats information and data security as a top priority, as detailed in its Trust Center and Terms of Service.
Accelerating Discovery with OpenConf on Tamarind Bio
Property prediction: feed diverse conformer ensembles into workflows such as pKa and logP calculations.
Flexible drug-like molecules: sample torsional space more completely than a seed-only approach.
Macrocycles: use specialized moves and an expanded energy window for ring systems.
Organometallic complexes: generate starting ensembles with restraints that keep metal-ligand bonds intact.
How to Use OpenConf on Tamarind Bio
Open the tool: log in to tamarind.bio and open OpenConf.
Provide the molecule: enter a SMILES string or upload a structure file, as the tool accepts.
Set the conformer budget: if exposed, choose how many conformers to generate; OpenConf is most advantageous at 50 or more.
Adjust options: if exposed, enable the macrocycle low-mode search or other settings.
Run the job: submit and wait for the ensemble.
Download outputs: retrieve the unique conformers.
Interpret: compare conformers by energy and diversity.
Use downstream: re-optimize and score conformers with higher-level methods.
Things to Keep in Mind
At low conformer counts (around 10) OpenConf is slower than ETKDG because of pool and seed overhead.
Macrocycles remain challenging even with specialized moves.
Performance on inorganic complexes is described as a bit flaky.
Benchmarking for some applications, such as logP, is early-stage.
The post does not state a license; check the repository.
Source: Wagen, Casetti, Vandezande and Mann, OpenConf and other open-source projects, Rowan newsletter, https://rowansci.substack.com/p/openconf-and-other-open-source-projects. Code: https://github.com/rowansci/openconf.