The Sonoma State University Department of Mathematics and Statistics presents Tao He, San Francisco State University.
T cells represent a crucial component of the adaptive immune system. Antigen-specific recognition is realized via T cell receptor (TCR), which is the product of somatic V(D)J gene recombination, plus some random addition/subtraction of nucleotides at recombination junctions. In this study, we aggregated the clones based on V/J gene segments, which overcomes the limitation and thus can build machine learning models across subjects. Here we presented a comparative study of different feature selection and classification on two multiclass Next-Generation Sequencing TCR repertoire data in cancer studies. We also proposed a novel ensemble of feature selection and demonstrated the method in simulation studies.
Series supported by Instructionally-Related Activities Funds. For more information about this series, go to www.sonoma.edu/math