{"success":true,"data":{"author":"Lawrence Livermore National Laboratory (Jones, Yang, Lightstone, Moshiri, Allen, Rosing)","bucketAccessType":"public","bucketName":"llnl/SurGBSA","bucketPrefix":"","category":"Drug Discovery","cloudProvider":"huggingface","createdAt":"2026-11-14T01:01:19.974Z","description":"Atomistic molecular dynamics (MD) trajectories for the PDBbind CASF-2016 benchmark, with snapshot-resolved Molecular Mechanics Generalized Born Surface Area (MM/GBSA) binding energy estimates, released by Lawrence Livermore National Laboratory with SurGBSA: Learning Representations From Molecular Dynamics Simulations (arXiv:2509.03084; LLNL-DATA-2016815). 10 ns Amber production simulations cover 242 of the 285 CASF-2016 protein-ligand complexes, each from up to 6 starting points: the crystal structure plus up to 5 AutoDock Vina docking poses. The release includes raw Amber MD files, ML-ready preprocessed arrays in NumPy format, per-snapshot MM/GBSA scores, k-fold cross-validation splits and trained model weights. Listing the repository (measured 2026-11-13) shows md.tar.gz (95.2 GB of raw trajectories), processed_data.tar.gz (4.45 GB), splits/ with 17 coreMD fold CSV files and weights/kfold/ with 15 EGMN model checkpoints; about 101 GB in all. Extract the archives and read trajectories with MDAnalysis, MDTraj or cpptraj, and the processed arrays with NumPy. Uses include surrogate models for binding free energy, protein-ligand affinity prediction and representation learning from MD. Released under CC-BY-4.0; cite the SurGBSA paper and PDBbind.","displayTitle":"SurGBSA: Molecular Dynamics Trajectories and MM/GBSA Scores for PDBbind CASF-2016","doi":"10.57967/hf/9379","id":"f19a8e0b-1480-5d5e-b4b1-36b79210bba2","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/llnl/SurGBSA/raw/main/README.md","objectCountEstimate":36,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2509.03084","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"llnl/SurGBSA\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"llnl/SurGBSA\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=llnl, license:cc-by-4.0, downloads=191"],"format":"mixed","generatedAt":"2026-11-14T01:01:24.666Z","layout":{"example":"splits/coreMD-fold-0-test.csv","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"3gnw\n4ea2\n1bcu\n3kgp\n1syi\n2wca\n4ih5\n2v00\n3jvr\n4ogj\n3ge7\n3ueu\n4qd6\n2wtv\n3f3c\n2yge\n1oyt\n3gr2\n4jsz\n2yfe\n3b68\n2cet\n4j28\n3uuo\n3bv9\n3n76\n4k18\n1k1i\n3lka\n3uex\n2xnb\n3up2\n3gc5\n2w66\n3acw\n5c2h\n3ryj\n1z9g\n3jya\n1bzc\n3zso\n2xdl\n3qgy\n1z6e\n3ao4\n3gbb\n1o3f\n2wer\n4twp\n","encoding":"utf8","key":"splits/coreMD-fold-0-test.csv","truncatedAt":512}],"sampleObjects":["md.tar.gz","processed_data.tar.gz","splits/coreMD-fold-0-test.csv","splits/coreMD-fold-0-train.csv","splits/coreMD-fold-0-val.csv"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\n---\n\n# Dataset Card for SurGBSA\n\n<!-- Provide a quick summary of the dataset. -->\n\n<!-- This dataset card aims to be a base template for new datasets. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1). -->\n\nData used in the development of *SurGBSA: Learning Representations From Molecular Dynamics Simulations*.\n\nJones, D., Yang, Y., Lightstone, F. C., Moshiri, N., Allen, J. E., & Rosing, T. S. (2025). SurGBSA: Learning representations from molecular dynamics simulations. In arXiv [q-bio.BM]. arXiv. http://arxiv.org/abs/2509.03084\n\n\nLLNL-DATA-2016815\n\n## Dataset Details\n\nAtomistic Molecular Dynamics trajectories for the PDBBind CASF-2016 benchmark set with snapshot-resolved Molecular Mechanics Generalized Born Surface Area measurements. \n\n\n### Dataset Description\n\n<!-- Provide a longer summary of what this dataset is. -->\n\nThis dataset is derived from the PDBBind CASF-2016 benchmark, as a representative subsample of the refined set. 10ns MD simulations are provided for 242/285 structures in the benchmark using up to 6 starting points per structure. The crystal structure is included as a starting point for all structures and up to 5 Vina docking poses are provided as additional starting points. We include \nthe raw Amber MD simulation files for the production runs, the preprocessed \"ml-ready\" files in numpy format, and mmgbsa scoring for each snapshot. We additionally provide the splits used for k-fold cross validation and the corresponding model weights. \n\n\n<!-- - **Curated by:** [More Information Needed] -->\n<!-- - **Funded by [optional]:** [More Information Needed] -->\n<!-- - **Funded by:**  -->\n<!-- - **Shared by [optional]:** [More Information Needed] -->\n<!-- - **Language(s) (NLP):** [More Information Needed] -->\n- **License:** Creative Commons Attribution 4.0 (CC-BY-4.0)\n\n### Dataset Sources\n\n<!-- Provide the basic links for the dataset. -->\n\n- **Repository:** https://github.com/llnl/SurGBSA\n- **Paper:** https://arxiv.org/abs/2509.03084\n<!-- - **Demo [optional]:** [More Information Needed] -->\n\n## Uses\n\n<!-- Address questions around how the dataset is intended to be used. -->\n\nThis resource is meant to be used for the development of machine learning algorithms and for applications in drug discovery research. \n\n<!-- ### Direct Use -->\n\n<!-- This section describes suitable use cases for the dataset. -->\n\n<!-- [More Information Needed] -->\n\n<!-- ### Out-of-Scope Use -->\n\n<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->\n\n<!-- [More Information Needed] -->\n\n<!-- ## Dataset Structure -->\n\n<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->\n\n<!-- [More Information Needed] -->\n\n<!-- ## Dataset Creation -->\n\n<!-- ### Curation Rationale -->\n\n<!-- Motivation for the creation of this dataset. -->\n\n<!-- [More Information Needed] -->\n\n### Source Data\n\nThe source data is taken from the [PDBBind database](https://www.pdbbind-plus.org.cn/).\n\n<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->\n\n<!-- #### Data Collection and Processing -->\n\n<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->\n\n<!-- [More Information Needed] -->\n\n<!-- #### Who are the source data producers? -->\n\n<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->\n\n<!-- [More Information Needed] -->\n\n<!-- ### Annotations [optional] -->\n\n<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->","encoding":"utf8","key":"README.md","name":"README.md","size":6917,"truncated":true}],"sourceObjectKey":"splits/coreMD-fold-0-test.csv","starterCell":"import pandas as pd\ndf = pd.read_csv(\"hf://datasets/llnl/SurGBSA/splits/coreMD-fold-0-test.csv\")\ndf.head()","starterSamplePath":"splits/coreMD-fold-0-test.csv","totalObjects":36},"summary":"Atomistic molecular dynamics (MD) trajectories for the PDBbind CASF-2016 benchmark, with snapshot-resolved Molecular Mechanics Generalized Born Surface Area (MM/GBSA) binding…","title":"SurGBSA: Molecular Dynamics Trajectories and MM/GBSA Scores for PDBbind CASF-2016","updatedAt":"2026-11-14T02:28:55.689Z","visibility":"public","websiteUrl":"https://github.com/llnl/SurGBSA"}}