遇见数据集

GEOM-DRUGS pre-simulated MD trajectories (Align Your Structures, ICLR 2026)

收藏
Zenodo2026-04-23 更新2026-05-26 收录
官方服务:

资源简介:

Pre-simulated molecular-dynamics trajectories for the GEOM-DRUGS subset used in Align Your Structures: Generating Trajectories with Structure Pretraining for Molecular Dynamics (ICLR 2026). Contents: 11,387 per-molecule simulation directories (5,682 train / 5,209 val / 496 test), corresponding to the 1,137 / 1,044 / 100 unique SMILES described in §5.2 of the paper. Each directory contains mol.pkl (RDKit Mol with hydrogens), system.pdb (OpenMM topology + initial positions), traj.xtc (5 ns trajectory at 4 fs timestep, frame_interval=100), and smiles.txt. Three *_mol_dirs.pkl split files under splits/ allow deterministic loader reproduction. Simulation settings: all-atom, explicit-solvent, 5 ns per trajectory, 4 fs timestep, hydrogen mass repartitioning at 1.5 amu, 100-step frame interval. Five independent replicas per SMILES. Multi-part archive: the 27 GB tarball is split into 5 GB parts for reliable upload. After downloading all parts, reassemble and extract in one command: cat geom-drugs-align-your-structures-v1.tar.gz.part* > geom-drugs-align-your-structures-v1.tar.gz tar xf geom-drugs-align-your-structures-v1.tar.gz Reference code: https://github.com/ani11452/Align_Your_Structures If you use this data please cite the paper above and the upstream GEOM dataset (Axelrod & Gómez-Bombarelli, Scientific Data 2022, doi:10.7910/DVN/JNGTDF).

提供机构:
Zenodo
创建时间:
2026-04-23
二维码
社区交流群
二维码
科研交流群
商业服务