AlgPred

AlgPred predicts allergenic proteins and maps IgE epitopes from protein sequences to assess allergenicity.

Key Features:

• Support Vector Machine (SVM): Utilizes amino acid and dipeptide composition for classification with reported accuracies of 85.02% and 84.00%, respectively.
• Motif-Based Methodology: Uses MEME/MAST to identify conserved motifs within allergenic proteins with reported sensitivity of 93.94% and specificity of 33.34%.
• Database Search for IgE Epitopes: Searches known IgE epitope databases to predict allergenicity with reported sensitivity of 17.47% and specificity of 98.14%.
• BLAST Search Against Allergen Peptides: Performs sequence similarity searches against a representative set of allergen peptides.
• Hybrid Approaches: Combines two or more methods to improve predictive performance by leveraging strengths of individual approaches.

Scientific Applications:

• Allergenic Protein Prediction: Assisting identification of potential allergens from novel protein sequences.
• IgE Epitope Mapping: Facilitating identification of epitopes that bind IgE antibodies to inform studies of allergic responses and design of hypoallergenic variants.
• Food Safety and Biotechnology: Supporting prediction of allergenic potential in genetically modified organisms (GMOs) and novel food products.

Methodology:

Data collection used training/testing sets of 578 allergens and 700 non-allergens and an independent validation set of 323 allergens and 101,725 non-allergens from Swiss-Prot; algorithms explicitly include SVM, MEME/MAST motif discovery, database searches for IgE epitopes, and BLAST sequence alignment, with performance assessed using independent datasets.