About AllergyPred

AllergyPred is a freely available webserver designed to predict potential food allergens, encompassing both protein and chemical-based allergens.

AllergyPred is an in silico tool designed for the prediction of food allergenicity in proteins and chemical compounds. It leverages advanced computational models, including machine learning, to assess the likelihood of a substance causing allergic reactions. The tool is particularly useful in the context of food safety, pharmaceutical product development, and regulatory assessments, as it minimizes reliance on traditional animal testing and supports quicker evaluations. Results are presented in a user-friendly table format, downloadable in several file formats, facilitating seamless integration into research or project reports.

Allergy Models

Computational Allergy Prediction

AllergyPred combines machine learning, bioinformatics, cheminformatics, statistics and immunology to assess potential allergenicity of proteins and chemical compounds such as food proteins, drugs, chemicals, or environmental agents.
AllergyPred uses labeled data on allergens and non-allergens to train respective classification models.
Using feature selection techniques, relevant features are extracted - such as amino acid composition, physicochemical properties, or epitope profiles. Algorithms includes support vector machine, deep neural network. It also provide a likelihood score for allergenicity.

Prediction Pipeline

Citation

Emanuel Kemmler, Emma Katherine Fath, Robert Preissner, Priyanka Banerjee. AllergyPred: a web server for allergen prediction, Nucleic Acids Res., May 6, 2025. PMID: 40326517, 10.1093/nar/gkaf383.