Scholarlink

An essential aspect of vaccine development and immunotherapy methods is the examination of how human leukocyte antigen (HLA) molecules convey antigens. This work presents a thorough analysis of computer models used to predict the binding affinities between peptides and HLA molecules. The models that are specifically examined include MARIA (Major Histocompatibility Complex Analysis with Recurrent Integrated Architecture), NNAlign, NetMHCIIpan, and PIA (Peptide Immune Annotation). These methods aim to find T-cell epitopes that interact with MHC-II molecules by integrating data from multiple experimental approaches, such as in-vitro binding tests and mass spectrometry (MS) sequencing. This integration of data helps in designing successful vaccinations and therapeutics. The analysis we conducted shows notable disparities in the performance of the model when applied to different datasets and evaluated using metrics such as Pearson correlation coefficient (PCC), Spearman correlation coefficient (SCC), and Area Under the Curve (AUC). MARIA exhibits exceptional predictive precision in detecting melanoma HLA-II-presented antigens and celiac-related gluten peptides, underscoring its promise in both cancer and autoimmune disease scenarios. This study highlights the significance of computational approaches in improving our comprehension of immune responses and antigen presentation, providing essential knowledge for the creation of targeted vaccinations and therapeutic therapies.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Review on HLA Class II Antigen Presentation Prediction via Microarray Assays, LC–MS, and Integrated Deep Learning

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Devanshi Sharma,
  • Nagendra Pratap Singh

摘要

An essential aspect of vaccine development and immunotherapy methods is the examination of how human leukocyte antigen (HLA) molecules convey antigens. This work presents a thorough analysis of computer models used to predict the binding affinities between peptides and HLA molecules. The models that are specifically examined include MARIA (Major Histocompatibility Complex Analysis with Recurrent Integrated Architecture), NNAlign, NetMHCIIpan, and PIA (Peptide Immune Annotation). These methods aim to find T-cell epitopes that interact with MHC-II molecules by integrating data from multiple experimental approaches, such as in-vitro binding tests and mass spectrometry (MS) sequencing. This integration of data helps in designing successful vaccinations and therapeutics. The analysis we conducted shows notable disparities in the performance of the model when applied to different datasets and evaluated using metrics such as Pearson correlation coefficient (PCC), Spearman correlation coefficient (SCC), and Area Under the Curve (AUC). MARIA exhibits exceptional predictive precision in detecting melanoma HLA-II-presented antigens and celiac-related gluten peptides, underscoring its promise in both cancer and autoimmune disease scenarios. This study highlights the significance of computational approaches in improving our comprehension of immune responses and antigen presentation, providing essential knowledge for the creation of targeted vaccinations and therapeutic therapies.