DeepImmuno Immunogenicity Prediction
DeepImmuno is a CNN-based model for peptide immunogenicity prediction with state-of-the-art accuracy across viral and cancer datasets.
DeepImmuno is a deep-learning model for accurate immunogenicity prediction of peptide–MHC complexes, achieving state-of-the-art performance across viral and tumor neoantigen datasets. Leveraging a convolutional neural network trained on a beta-binomial scoring framework, it assigns continuous immunogenicity scores that reflect experimental confidence, outperforming traditional classifiers and existing tools like IEDB and DeepHLApan. DeepImmuno-CNN integrates physicochemical-aware amino acid encodings to model TCR–peptide–MHC interactions and systematically identifies the most salient residues for antigen recognition. Designed for broad HLA allele coverage and stable across varied dataset sizes, the model delivers reliable prioritization of immunogenic epitopes for vaccine and immunotherapy development.
Estimated resource cost: 0.000062
Categories: Protein Design
Tags: Developability Immunogenicity Proteins
Key Capabilities
- Beta-binomial scoring layer outputs continuous immunogenicity scores with calibrated confidence.
- CNN-based architecture tuned for peptide–MHC interaction modeling across diverse HLA alleles.
- Saliency-mapped residue importance highlights dominant TCR-facing positions (P4–P6).
- Systematic benchmarking across dengue, cancer neoantigen, and SARS-CoV-2 datasets.
- Independent of MHC-binding models; complements existing HLA-binding tools in immunotherapy pipelines.
Runtime Statistics
| Metric | Value |
|---|---|
| runtime_mean | 4 |
| runtime_median | 3 |
| runtime_std | 2 |
| runtime_90th_percentile | 9 |
| runtime_max | 28 |
Similar Tools
- BindFilter
- BioPhi
- Humatch
- BioBind
Ready to submit your job?
Review your configuration, then confirm the estimated credit cost before you run the job. Note that credit estimates are not guaranteed and runtime can vary depending on inputs and settings.
Estimated Credits: 0.000062
Invite-only, limited-time access. Please contact ztang@getantibody.com.
Invite-only, limited-time access. Please contact ztang@getantibody.com.