Antibody Specificity Predictor (ASPred)
ASPred is an AI framework that identifies antigen-specific B-cell receptor sequences directly from immune repertoires.
It was created to help researchers rapidly detect which antibodies in a sample are likely to recognize a particular viral antigen, without requiring paired chains or structural modeling.
ASPred learns antigen-specific patterns from experimentally validated datasets and provides pretrained models for three major viral targets:
These models allow users to screen bulk or single-cell BCR repertoires and estimate which sequences are enriched for antigen recognition. ASPred can be applied for antibody discovery, repertoire characterization, immune monitoring, and hypothesis generation in vaccine and infectious-disease research.
By making these three pretrained classifiers publicly available, ASPred offers a simple and scalable tool for exploring antigen specificity across diverse immune repertoires.