| | --- |
| | license: cc-by-4.0 |
| | tags: |
| | - biology |
| | - chemistry |
| | - medical |
| | - immunology |
| | - immunogenicity |
| | pretty_name: ImmunoStruct |
| | size_categories: |
| | - 10K<n<100K |
| | --- |
| | |
| | # ImmunoStruct |
| | [ImmunoStruct enables multimodal deep learning for immunogenicity prediction](https://www.nature.com/articles/s42256-025-01163-y) |
| |
|
| | <div align="left"> |
| | |
| | [](https://www.nature.com/articles/s42256-025-01163-y) |
| | [](https://www.nature.com/articles/s42256-025-01163-y.pdf) |
| |  |
| | [](https://huggingface.co/datasets/ChenLiu1996/ImmunoStruct) |
| | [](https://huggingface.co/ChenLiu1996/ImmunoStruct) |
| | [](https://github.com/KrishnaswamyLab/ImmunoStruct) |
| | <br>[](https://www.linkedin.com/in/kevin-bijan-givechian-phd-36467ba3/) |
| | [](https://www.linkedin.com/in/joao-felipe-rocha/) |
| | [](https://www.linkedin.com/in/chenliu1996/) |
| | [](https://scholar.google.com/citations?user=3rDjnykAAAAJ&sortby=pubdate) |
| | <br>[](https://x.com/KevinGivechian) |
| | [](https://x.com/ChenLiu_1996) |
| | [](https://x.com/KrishnaswamyLab) |
| | </div> |
| |
|
| | Project leads: [Kevin Bijan Givechian](https://www.linkedin.com/in/kevin-bijan-givechian-phd-36467ba3/), [João Felipe Rocha](https://www.linkedin.com/in/joao-felipe-rocha/), [Chen Liu](https://www.linkedin.com/in/chenliu1996/). |
| | <br>Correspondence: `akiko.iwasaki@yale.edu`, `smita.krishnaswamy@yale.edu`. |
| |
|
| | Instructions on preparing everything and running training/inference is provided on [our official GitHub repository](https://github.com/KrishnaswamyLab/ImmunoStruct). |
| |
|
| | The pre-trained model weights for IEDB and CEDAR datasets are available at [our huggingface model repo](https://huggingface.co/ChenLiu1996/ImmunoStruct). |
| |
|
| | In short, the data can be downloaded using |
| | <br>`hf download ChenLiu1996/ImmunoStruct --repo-type dataset --local-dir ./` |
| |
|
| | ## Dataset details |
| |
|
| | In this huggingface dataset, we include all data used in the paper. |
| |
|
| | - Necessary for running training and/or inference on IEDB: 1, 2. |
| | - Necessary for running training and/or inference on CEDAR: 1, 2. |
| | - Necessary for running inference on clinical validation data: 1, 2. |
| | - Necessary if you want to build your graph differently: 3. |
| |
|
| | 1. CSV files of (protein sequences, biochemical property values, and immunogenicity scores) for all 3 datasets (IEDB, CEDAR, and clinical validation), CSV file of clinical survival data, and CSV file of MHC (a.k.a. HLA) sequences. |
| | ``` |
| | ImmunoStruct_IEDB_data.csv |
| | ImmunoStruct_CEDAR_data_cancer.csv |
| | ImmunoStruct_CEDAR_data_wildtype.csv |
| | ImmunoStruct_clinical_data.csv |
| | ImmunoStruct_clinical_data_survival.csv |
| | HLA_allele_sequences.csv |
| | ``` |
| | 2. AlphaFold2 structures, in PyTorch Geometric format. |
| | ``` |
| | graph_pyg_IEDB.zip |
| | graph_pyg_CEDAR_cancer.zip |
| | graph_pyg_CEDAR_wildtype.zip |
| | graph_pyg_clinical.zip |
| | ``` |
| | 3. (Optional) AlphaFold2 structures, in raw PDB format. |
| | ``` |
| | alphafold2_pdb_IEDB.zip |
| | alphafold2_pdb_CEDAR_cancer.zip |
| | alphafold2_pdb_CEDAR_wildtype.zip |
| | alphafold2_pdb_clinical.zip |
| | ``` |
| |
|
| | ## Citation |
| |
|
| | If you use ImmunoStruct in your research, please cite our paper: |
| |
|
| | BibTeX: |
| | ```bibtex |
| | @article{givechian2026immunostruct, |
| | title={ImmunoStruct enables multimodal deep learning for immunogenicity prediction}, |
| | author={Givechian, Kevin Bijan and Rocha, Jo{\~a}o Felipe and Liu, Chen and Yang, Edward and Tyagi, Sidharth and Greene, Kerrie and Ying, Rex and Caron, Etienne and Iwasaki, Akiko and Krishnaswamy, Smita}, |
| | journal={Nature Machine Intelligence}, |
| | volume={8}, |
| | pages={70--83}, |
| | year={2026}, |
| | publisher={Nature Publishing Group UK London} |
| | } |
| | ``` |
| | Nature format:<br> |
| | Givechian, K.B., Rocha, J.F., Liu, C. et al. ImmunoStruct enables multimodal deep learning for immunogenicity prediction. *Nat Mach Intell* 8, 70–83 (2026). https://doi.org/10.1038/s42256-025-01163-y |
| |
|
