{"id":18983120,"url":"https://github.com/rosettacommons/deepab","last_synced_at":"2025-04-15T07:06:31.052Z","repository":{"id":42046993,"uuid":"371386193","full_name":"RosettaCommons/DeepAb","owner":"RosettaCommons","description":"Deep learning models and structure realization scripts for the DeepAb antibody structure prediction method.","archived":false,"fork":false,"pushed_at":"2022-11-30T20:18:36.000Z","size":206,"stargazers_count":158,"open_issues_count":12,"forks_count":31,"subscribers_count":13,"default_branch":"main","last_synced_at":"2025-04-15T07:06:28.539Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/RosettaCommons.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.md","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2021-05-27T13:37:59.000Z","updated_at":"2025-03-31T07:49:45.000Z","dependencies_parsed_at":"2022-08-29T18:30:32.301Z","dependency_job_id":null,"html_url":"https://github.com/RosettaCommons/DeepAb","commit_stats":null,"previous_names":[],"tags_count":3,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RosettaCommons%2FDeepAb","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RosettaCommons%2FDeepAb/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RosettaCommons%2FDeepAb/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/RosettaCommons%2FDeepAb/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/RosettaCommons","download_url":"https://codeload.github.com/RosettaCommons/DeepAb/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":249023700,"owners_count":21199958,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-11-08T16:15:58.480Z","updated_at":"2025-04-15T07:06:31.016Z","avatar_url":"https://github.com/RosettaCommons.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# _DeepAb_\nOfficial repository for [DeepAb](https://www.sciencedirect.com/science/article/pii/S2666389921002804): Antibody structure prediction using interpretable deep learning.  The code, data, and weights for this work are made available under the [Rosetta-DL license](LICENSE.md) as part of the [Rosetta-DL](https://github.com/RosettaCommons/Rosetta-DL) bundle.\n\nTry antibody structure prediction in [Google Colab](https://colab.research.google.com/github/RosettaCommons/DeepAb/blob/main/DeepAb.ipynb).\n\n## Setup\n\n_Optional_: Create and activate a python virtual environment\n```\npython3 -m venv venv\nsource venv/bin/activate\n```\nInstall project dependencies\n```\npip install -r requirements.txt\n```\n\n_Note_: PyRosetta should be installed following the instructions [here](http://pyrosetta.org/downloads).\n\nDownload pretrained model weights\n```\nwget https://data.graylab.jhu.edu/ensemble_abresnet_v1.tar.gz\ntar -xf ensemble_abresnet_v1.tar.gz\n```\nAfter unzipping, pre-trained models might need to be moved such that they have paths `trained_models/ensemble_abresnet/rs*.pt`\n\n## Common workflows\n\nAdditional options for all scripts are available by running with `--help`.\n\n_Note_: This project is tested with Python 3.7.9\n\n_Note_: Using `--renumber` option will send your antibody to the [AbNum server](http://www.bioinf.org.uk/abs/abnum/). If working with confidential sequences you should avoid this option and use an external renumbering tool.\n\n\n### Structure prediction\nGenerate an antibody structure prediction from an Fv sequence with five decoys:\n```\npython predict.py data/sample_files/4h0h.fasta --decoys 5 --renumber\n```\nGenerate a structure for a single heavy or light chain:\n```\npython predict.py data/sample_files/4h0h.fasta --decoys 5 --single_chain\n```\n_Note_: The fasta file should contain a single entry labeled \"H\" (even if the sequence is a light chain).\n\n**Expected output**\n\nAfter the script completes, the final prediction will be saved as `pred.deepab.pdb`.  The numbered decoy structures will be stored in the `decoys/` directory.\n\n\n### Attention annotation\nAnnotate an Fv structure with H3 attention:\n```\npython annotate_attention.py data/sample_files/4h0h.truncated.pdb --renumber --cdr_loop h3\n```\n_Note_: CDR loop residues are determined using Chothia definitions, so the input structure should be numbered beforehand or renumbered by passing `--renumber`\n\n**Expected output**\n\nAfter the script completes, the annotated PDB will overwrite the input file (unless `--out_file` is specificed).  Annotations will be stored as b-factor information, and can be visualized in [PyMOL](https://pymol.org/2/) or similar software.\n\n### Design scoring\nCalculate ΔCCE for list of designed sequences:\n```\npython score_design.py data/sample_files/wt.fasta data/sample_files/h_mut_seqs.fasta data/sample_files/l_mut_seqs.fasta design_out.csv\n```\n\n**Expected output**\n\nAfter the script completes, the designs and scores will be written to a CSV file with each row containing the design ID, heavy chain sequence, light chain sequence, and  ΔCCE value.\n\n## References\n[1] JA Ruffolo, J Sulam, and JJ Gray. \"[Antibody structure prediction using interpretable deep learning.](https://www.sciencedirect.com/science/article/pii/S2666389921002804)\" _Patterns_ (2022).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frosettacommons%2Fdeepab","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frosettacommons%2Fdeepab","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frosettacommons%2Fdeepab/lists"}