A Python package for easy molecular docking with a curated dataset and benchmark tasks for drug discovery.
dockstring is a Python package that simplifies molecular docking calculations for drug discovery research. It provides an easy-to-use interface for docking molecules from SMILES strings and includes a curated dataset with benchmark tasks to evaluate ligand design methods. The package helps researchers perform reproducible docking experiments and benchmark their algorithms against standardized tasks.
Computational chemists, bioinformaticians, and drug discovery researchers who need to perform molecular docking calculations and benchmark ligand design algorithms. It's particularly useful for those working on machine learning for drug discovery who require standardized evaluation metrics.
dockstring offers a streamlined Python interface for molecular docking combined with a carefully curated dataset, making it easier to perform reproducible docking experiments and benchmark ligand design methods. Unlike individual docking tools, it provides standardized benchmark tasks and pre-computed scores specifically designed for evaluating drug discovery algorithms.
A Python package for molecular docking with an extensive, highly-curated dataset and a set of realistic benchmark tasks for drug discovery.
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Allows docking from SMILES strings in just a few lines of Python code, as demonstrated in the tutorials, making it accessible for rapid prototyping.
Includes a pre-computed dataset with docking scores for standardized evaluation of ligand design algorithms, ensuring reproducibility as cited in the paper.
Compatible with PyMol for visualizing targets, search boxes, and ligands, enhancing interpretability with easy setup via conda.
Provides pytest scripts to verify installation matches the dataset, supporting reliable benchmarking with clear error rate reporting.
Requires conda for installation because openbabel cannot be installed via pip, adding complexity for environments that rely solely on pip.
Mac support is limited, with scores not always matching Linux versions, which can undermine cross-platform reproducibility in research.
To achieve exact dataset matching, specific package versions like python 3.7 are needed, but this version is end-of-life, complicating modern deployments.