Pcleavage

Pcleavage predicts proteasome cleavage sites within antigenic sequences to inform antigen processing and major histocompatibility complex (MHC) peptide presentation analyses.

Key Features:

• Algorithmic Foundation: Employs a support vector machine (SVM) trained to distinguish cleavage patterns processed by constitutive proteasomes and immunoproteasomes.
• Performance Metrics: Reports Matthew's correlation coefficients (MCC) of 0.54 for in vitro data and 0.43 for major histocompatibility complex ligand data, performing comparably to the NetChop method.

Scientific Applications:

• Antigen processing and presentation: Predicts proteasomal cleavage events to support analysis of peptide generation for presentation by MHC molecules to T cells.
• Epitope identification: Supports selection of candidate epitopes for vaccine development by identifying likely proteasomal cleavage sites within proteins.
• Cancer immunotherapy research: Assists identification of tumor-derived peptides produced by proteasomal cleavage for neoantigen discovery.

Methodology:

An SVM model was trained on datasets of known proteasome cleavage sites to learn patterns distinguishing constitutive and immunoproteasomal processing and to enable prediction in novel sequences.