RosettaCommons/SAbDab_raw

1---2language:3- en4license: cc-by-4.05license_link: LICENSE.md6tags:7- biology8- chemistry9dataset_summary: All data available on The Structural Antibody Database (SAbDab)10pretty_name: Raw data from The Structural Antibody Database (SAbDab)11dataset_description: This dataset contains all available data from the SAbDab as of Mar 4, 2026. The SAbDab is a database of antibody structures  including experimental details, antibody nomenclature, affinity data and sequence annotations.12acknowledgements: 'We kindly acknowledge the SAbDab team, RosettaCommons, and the following institutions: University of California, Los Angeles; University of Maryland; University of Oregon; University of Michigan; University of Pennsylvania; and the Wistar Institute'13size_categories:14- 10K<n<100K15citation_bibtex: >-16  @article{10.1093/nar/gkt1043, 17  author = {Dunbar, James and Krawczyk, Konrad and Leem, Jinwoo and 18  Baker, Terry and Fuchs, Angelika and Georges, Guy and Shi, Jiye 19  and Deane, Charlotte M.}, title = {SAbDab: the structural antibody 20  database}, journal = {Nucleic Acids Research}, volume = {42}, 21  number = {D1}, pages = {D1140-D1146}, year = {2013}, month = {11},22  abstract = {Structural antibody database (SAbDab; http://opig.stats.ox.ac.uk/webapps/sabdab)23  is an online resource containing all the publicly available antibody24  structures annotated and presented in a consistent fashion. The data25  are annotated with several properties including experimental information,26  gene details, correct heavy and light chain pairings, antigen details and,27  where available, antibody–antigen binding affinity. The user can select 28  structures, according to these attributes as well as structural properties29  such as complementarity determining region loop conformation and variable30  domain orientation. Individual structures, datasets and the complete database31  can be downloaded.}, issn = {0305-1048}, doi = {10.1093/nar/gkt1043},32  url = {https://doi.org/10.1093/nar/gkt1043},33  eprint = {https://academic.oup.com/nar/article-pdf/42/D1/D1140/3538157/gkt1043.pdf}}34citation_apa: >-35  James Dunbar, Konrad Krawczyk, Jinwoo Leem, Terry Baker, Angelika Fuchs, Guy Georges, 36  Jiye Shi, Charlotte M. Deane, SAbDab: the structural antibody database, Nucleic Acids 37  Research, Volume 42, Issue D1, 1 January 2014, Pages D1140–D1146, 38  https://doi.org/10.1093/nar/gkt104339---40 41# All raw data from The Structural Antibody Database (SAbDab) 42 43## Quickstart Usage44 45### Install HuggingFace Datasets package46 47Each subset can be loaded into python using the Huggingface [datasets](https://huggingface.co/docs/datasets/index) library.48First, from the command line install the `datasets` library49 50    $ pip install datasets51 52Optionally set the cache directory, e.g.53 54    $ HF_HOME=${HOME}/.cache/huggingface/55    $ export HF_HOME56 57then, from within python load the datasets library58 59    >>> import datasets60 61### Load model datasets62   63To load structures from the entire `SAbDab` dataset, use `datasets.load_dataset(...)`:64 65    >>> dataset_tag = "train"66    >>> dataset_models = datasets.load_dataset(67      path = "ProteinMPNN/SAbDab_raw",68      name = f"{dataset_tag}_models",69      data_dir = f"{dataset_tag}")['train']70 71and the dataset is loaded as a `datasets.arrow_dataset.Dataset`72 73    >>> dataset_models74    Dataset({75        features: [76          'pdb',77          'Hchain',78          'Lchain',79          'model',80          'antigen_chain',81          'antigen_type',82          'antigen_het_name',83          'antigen_name',84          'short_header',85          'date',86          'compound',87          'organism',88          'heavy_species',89          'light_species',90          'antigen_species',91          'authors',92          'resolution',93          'method',94          'r_free',95          'r_factor',96          'scfv',97          'engineered',98          'heavy_subclass',99          'light_subclass',100          'light_ctype',101          'affinity',102          'delta_g',103          'affinity_method',104          'temperature',105          'pmid'106          ],107        num_rows: 20701108    })109 110which is a column oriented format that can be accessed directly, converted in to a `pandas.DataFrame`, or `parquet` format, e.g.111 112    >>> dataset_models.data.column('pdb')113    >>> dataset_models.to_pandas()114    >>> dataset_models.to_parquet("dataset.parquet")115 116## Dataset Details117 118### Dataset Description119This dataset contains all available data from the SAbDab as of Mar 4, 2026. The SAbDab is a database of antibody structures  including experimental details, antibody nomenclature, affinity data and sequence annotations.120 121- **Acknowledgements:**122  We kindly acknowledge the SAbDab team, RosettaCommons, and the following institutions: University of California, Los Angeles; University of Maryland; University of Oregon; University of Michigan; University of Pennsylvania; and the Wistar Institute.123 124- **License:** CC-BY 4.0125 126### Dataset Sources127- **Paper:** Dunbar, J., Krawczyk, K. et al (2014). Nucleic Acids Res. 42. D1140-D1146128 129## Uses130Screening of antibody-antigen interactions, querying structural features of antibodies, and more131 132## Citation133  @article{10.1093/nar/gkt1043, 134  author = {Dunbar, James and Krawczyk, Konrad and Leem, Jinwoo and Baker, Terry and Fuchs, Angelika and Georges, Guy and Shi, Jiye and Deane, Charlotte M.},135  title = {SAbDab: the structural antibody database},136  journal = {Nucleic Acids Research}, 137  volume = {42}, 138  number = {D1},139  pages = {D1140-D1146},140  year = {2013},141  month = {11},142  abstract = {Structural antibody database (SAbDab; http://opig.stats.ox.ac.uk/webapps/sabdab) is an online resource containing all the publicly available antibody structures annotated and presented in a consistent fashion. The data are annotated with several properties including experimental information, gene details, correct heavy and light chain pairings, antigen details and, where available, antibody–antigen binding affinity. The user can select structures, according to these attributes as well as structural properties such as complementarity determining region loop conformation and variable domain orientation. Individual structures, datasets and the complete database can be downloaded.},143  issn = {0305-1048},144  doi = {10.1093/nar/gkt1043},145  url = {https://doi.org/10.1093/nar/gkt1043},146  eprint = {https://academic.oup.com/nar/article-pdf/42/D1/D1140/3538157/gkt1043.pdf}147  }148 149## Dataset Card Authors150Miranda Simpson (miranda13nicoles@gmail.com), Becca Lee (beccalee5@g.ucla.edu), Nathaniel Felbinger (nfelbing@umd.edu), Pratyush Dhal (pdhal@umich.edu), Colby Agostino (colby.agostino@pennmedicine.upenn.edu)