Towards a Mathematical Foundation of Immunology and Amino Acid Chains

摘要: We attempt to set a mathematical foundation of immunology and amino acid chains. To measure the similarities these chains, kernel on strings is defined using only sequence chains good substitution matrix (e.g. BLOSUM62). The used in learning machines predict binding affinities peptides human leukocyte antigen DR (HLA-DR) molecules. On both fixed allele [24] pan-allele [23] benchmark databases, our algorithm achieves state-of-the-art performance. also define distance an HLA-DR based which clustering analysis precisely recovers serotype classifications assigned by WHO [14, 22]. These results suggest that relates well chain structure molecules their biological functions, it offers simple, powerful promising methodology studies.