NetMHCIIphosPan: HLA Class II Phosphorylated Peptide Presentation Predictions

NetMHCIIphosPan is a machine learning tool for predicting antigen presentation of phosphorylated peptides by human leukocyte antigen (HLA) class II molecules. The method was developed by Heli M. Garcia Alvarez using mass spectrometry-based immunopeptidomics datasets and was last updated in April 2026. It demonstrates superior performance compared to existing tools like NetMHCIIpan-4.3 and MixMHC2pred-1.3.

Use Cases

• Predict HLA class II antigen presentation of phosphorylated peptides based on mass spectrometry immunopeptidomics data.
• Analyze HLA-specific binding motifs for phospholigands compared to unmodified ligands.
• Train or benchmark predictive models for HLA-DP and HLA-DQ molecule specificity.
• Support research into immune regulation and the HLA-II phospholigandome for therapeutic applications.

Strengths

• The tool is described as achieving superior performance compared to existing state-of-the-art tools NetMHCIIpan-4.3 and MixMHC2pred-1.3.
• Predictive accuracy was enhanced by incorporating unmodified ligands into training, particularly for HLA-DP and HLA-DQ molecules.
• The dataset is small at 56.0 KB, facilitating quick download and initial inspection.

Limitations

• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is tiny (56.0 KB), indicating limited scope, likely containing model outputs or parameters rather than extensive raw data.

Provenance

Source

Heli M. Garcia Alvarez via figshare.

Collection Method

Developed using mass spectrometry (MS)-based immunopeptidomics datasets and a refined peptide identification workflow.

Time Range

null

Freshness

Last updated 2026-04-17 09:29:50; freshness should be verified.

Geography

null