Related Experiment Video
Updated: Dec 17, 2025
09:28
Enrich and Expand Rare Antigen-specific T Cells with Magnetic Nanoparticles
Published on: November 17, 2018
11.9K
DeepAntigen: a novel method for neoantigen prioritization via 3D genome and deep sparse learning
Yi Shi1,2,3, Zehua Guo2,4, Xianbin Su1
1Key Laboratory of Systems Biomedicine (Ministry of Education), Shanghai Centre for Systems Biomedicine, Shanghai Jiao Tong University, Shanghai 200240, China.
Bioinformatics (Oxford, England)
|June 28, 2020
Summary
Cancer mutations can create neoantigens for T-cell recognition. We developed a deep learning model incorporating 3D genome data to improve neoantigen prioritization for cancer immunotherapy.
Area of Science:
- Oncology
- Immunology
- Bioinformatics
Background:
- Cancer mutations can generate neoantigens, which are targets for T-cell immunotherapy.
- Accurately prioritizing neoantigens is computationally challenging, especially with abundant mutations.
- Existing methods lack advanced machine learning models and do not consider 3D genome structure for neoantigen immunogenicity prediction.
Purpose of the Study:
- To develop an advanced machine learning model for accurate neoantigen prioritization.
- To investigate the role of 3D genome structure in neoantigen immunogenicity.
- To provide a novel computational tool for personalized cancer immunotherapy.
Main Methods:
- Discovered distinct spatial distribution patterns of immunopositive and immunonegative MHC-I neoantigens in the 3D genome.
- Developed a deep sparse neural network model with group feature selection (DNN-GFS).
- Integrated 3D genome information and an ensemble pMHC-I coding strategy into the DNN-GFS model.
Main Results:
- The DNN-GFS model demonstrated superior neoantigen prioritization power compared to existing sequence-based methods.
- Distinct 3D genome spatial distribution patterns were identified for neoantigens.
- A webserver, deepAntigen, was developed to implement the DNN-GFS model and other machine learning methods.
Conclusions:
- The study introduces a novel perspective on neoantigen prediction by incorporating 3D genome information.
- The DNN-GFS model offers improved accuracy in prioritizing neoantigens.
- The deepAntigen webserver provides a valuable tool for advancing personalized cancer immunotherapy research.
