DTU Health Tech

Department of Health Technology


ICORE-based prediction of neo-epitope immunogenicity.

ICERFIRE is an ensemble of Random Forest models. It predicts neo-epitope immunogenicity by identifying the best HLA-binding ICORE and uses self-similarity, a mutation score, and antigen expression.

Submit data

Paste in the sequences for the mutant, wild-type, and a third column containing the HLA alleles (in the format HLA-XYYZZ, ex: HLA-A0201). Sequences should be 8 to 14 amino acids long.
Additionally, the expression values of each data point can be provided by the user, if available and the option below is selected. If no expression values are provided and expression is included in the model, expression values will be queried from a reference dataset (TCGA pan-can).
Alternatively, load and example input or upload a file from your local machine.
Each column should be comma separated.

For detailed instructions, see Instructions tab above.

For an overview of the method and citation information, see Abstract tab.

Sequence submission

Paste the sequence(s):

or load some sample data:
or upload a local file:


Include expression in the model (Recommended): Yes   No
Use user-provided expression values: Yes   No


Instructions for ICERFIRE-1.0

Input format

  • The data should have format: mutant,wild-type,HLA, or mutant,wild-type,HLA,TPMs.
  • The server only accepts peptidic sequences of length 8 to 14, with the standard uppercase 20 amino acid alphabet:

  • A C D E F G H I K L M N P Q R S T V W Y
  • The HLA alleles must be provided with full resolution, in the format "HLA-XYYZZ", ex: HLA-A0201
  • The user can provide their own expression values if available. Else, the model will first query reference expression values from the TCGA pan-can dataset.
  • If the user chooses to provide their own expression values, the option "Use user-provided expression values" must be checked.
  • The input needs to be comma separated, with no headers.

  • Load Example data on the Submission page to ensure your data format is correct.


  1. Paste the data into the box, or load an example file, or load a file from your local machine. If "Use user-provided expression values" is not checked, 3 columns are expected in the input file; if it is checked, 4 columns are expected.
  2. Select whether to include expression in the model. It is not recommended to run the model without expression values, even if no TPM values are provided; the model will automatically query reference expression values.
  3. Select whether to provide your own TPM values.
Click the submit button when all data are entered.


A large-scale study of peptide features defining immunogenicity of cancer neo-epitopes Yat-tsai Richie Wan, Zeynep Kosalogly-Yalcon, Bjoern Peters, Morten Nielsen


Accurate prediction of immunogenicity for neo-epitopes is crucial for the development of personalised cancer immunotherapies and vaccines. In this study, we performed a comprehensive study of peptide features relevant for prediction of immunogenicity using the Cancer Epitope Database and Analysis Resource (CEDAR), a curated database of cancer epitopes with experimentally validated immunogenicity annotations from peer-reviewed publications.
The developed model, ICERFIRE (ICore-based Ensemble Random Forest for neo-epitope Immunogenicity pREdiction), takes as input the predicted ICORE rather than the full neopeptide as input, i.e. the submer with the highest predicted major histocompatibility complex (MHC) binding potential combined with its predicted likelihood of antigen presentation (%Rank).
Key additional features integrated into the model include self-similarity score, the similarity to the aligned wild-type ICORE to assess self-tolerance; BLOSUM mutation score, scoring a mutant against its wild-type counterpart; and wild-type antigen expression, capturing a neo-epitope’s abundance. We demonstrate improved and robust performance of ICERFIRE over existing immunogenicity and epitope prediction models, both in cross-validation and on external validation datasets.

Software Downloads

  • Version 1.0a


If you need help regarding technical issues (e.g. errors or missing results) contact Technical Support. Please include the name of the service and version (e.g. NetPhos-4.0) and the options you have selected. If the error occurs after the job has started running, please include the JOB ID (the long code that you see while the job is running).

If you have scientific questions (e.g. how the method works or how to interpret results), contact Correspondence.

Correspondence: Technical Support: