ImmunoStruct

ImmunoStruct enables multimodal deep learning for immunogenicity prediction

Project leads: Kevin Bijan Givechian, João Felipe Rocha, Chen Liu.
Correspondence: akiko.iwasaki@yale.edu, smita.krishnaswamy@yale.edu.

Instructions on preparing everything and running training/inference is provided on our official GitHub repository.

The pre-trained model weights for IEDB and CEDAR datasets are available at our huggingface model repo.

In short, the data can be downloaded using
hf download ChenLiu1996/ImmunoStruct --repo-type dataset --local-dir ./

NOTE: Our official repository also includes detailed instructions and working code on how to run protein folding using AlphaFold2.

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 find ImmunoStruct helpful to your research, please cite our paper:

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:
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