PyGeoDock is a Python prototype for geometric molecular expansion in molecular docking. This project is inspired by the paper "Tunable Approximations to Control Time-to-Solution in an HPC Molecular Docking Mini-App".
This code serves as the basis for the research published in:
Please note that the original code of the publication is now the intellectual property of Dompe Pharmaceutics.
PyGeoDock is a simplified version of a full-scale molecular docking application. This Mini-App is used for:
- Testing and validating the approach
- Easy experimentation and benchmarking
The Mini-App is designed with modularity in mind, allowing for easy tweaking of knobs and integration with the autotuning system.
PyGeoDock is optimized for High-Performance Computing systems, featuring:
- Parallel execution optimization
- Efficient memory usage
- Scalability across multiple processors
The code incorporates:
- Parallelization techniques (e.g., using MPI or OpenMP)
- Optimized data structures
- Efficient I/O operations
- MPI4PY, Multiprocessing