Related Experiment Video
Updated: Jul 3, 2025
09:28
Enrich and Expand Rare Antigen-specific T Cells with Magnetic Nanoparticles
Published on: November 17, 2018
11.6K
ImmuneMirror: A machine learning-based integrative pipeline and web server for neoantigen prediction
Gulam Sarwar Chuwdhury1, Yunshan Guo2, Chi-Leung Chiang1
1Department of Clinical Oncology, Center of Cancer Medicine, School of Clinical Medicine, Li Ka Shing Faculty of Medicine, University of Hong Kong, Hong Kong (SAR), P. R. China.
Briefings in Bioinformatics
|February 12, 2024
Summary
Neoantigens, derived from tumor mutations, can trigger anti-tumor immune responses. ImmuneMirror, a new tool, accurately predicts neoantigens and identifies patient subgroups who may not respond to PD-1 blockade therapy.
Area of Science:
- Immunology
- Bioinformatics
- Oncology
Background:
- Neoantigens arise from somatic mutations and can elicit T-cell responses against tumors.
- Identifying neoantigens is crucial for developing targeted cancer immunotherapies.
- Existing neoantigen prediction methods require comprehensive evaluation.
Purpose of the Study:
- To develop and validate ImmuneMirror, an open-source pipeline and web server for neoantigen prediction and prioritization.
- To apply ImmuneMirror to gastrointestinal cancer data to identify novel therapeutic targets and patient subgroups.
- To assess the utility of neoantigen load in predicting response to PD-1 blockade in microsatellite instability-high colorectal cancer patients.
Main Methods:
- Developed a balanced random forest model for neoantigen prediction, trained on experimentally validated neopeptides.
- Validated the model using independent testing data, achieving an area under the curve of 0.87.
- Applied ImmuneMirror to whole-exome and RNA sequencing data from 805 gastrointestinal cancer patients (CRC, ESCC, HCC).
Main Results:
- Identified a subgroup of microsatellite instability-high (MSI-H) colorectal cancer (CRC) patients with high tumor mutation burden but low neoantigen load.
- This subgroup showed significantly lower neoantigen presentation for MHC class I and II molecules, suggesting potential non-response to PD-1 blockade.
- Discovered a specific actionable neopeptide (YMCNSSCMGV-TP53G245V) in esophageal squamous cell carcinoma (ESCC) restricted by HLA-A02.
Conclusions:
- ImmuneMirror is a reliable and effective tool for neoantigen prediction and prioritization.
- Neoantigen load can help stratify MSI-H CRC patients for PD-1 blockade therapy.
- The study provides valuable insights into neoantigen landscapes in gastrointestinal cancers and identifies potential therapeutic targets.
