A next-gen pipeline for generation, error correction and annotation of high-throughput immunosequencing data (TECH1P.842)

Abstract

Abstract Immunosequencing is emerging as a new technology with substantial scientific and clinical relevance in immunology, hematology and oncology. Thanks to the highly-randomized somatic rearrangement of antigen receptor genes in lymphocytes, each clonal population of lymphocytes bears a receptor rearrangement that is essentially unique. Thus, sequencing of rearranged antigen receptors in B and T cells can provide a thorough characterization of the diversity and clonal structure of the adaptive immune system. Despite the significant promise of immunosequencing, accurate and consistent methods for data generation and cleaning have proven elusive in practice. Outstanding issues include biased amplification of rearranged antigen receptor genes; computational correction of PCR and sequencing errors in raw sequencing data; alignment of receptor sequences to germline gene receptor loci for accurate reconstruction of rearrangement events; and methods for extrapolation from sequencing reads attributed to each unique rearrangement to absolute input cell numbers. Using a set of robust experiments in which carefully-controlled conditions and synthetic molecules are utilized to precisely determine correct output, we have developed an experimental paradigm and data analysis pipeline that address these concerns and set a new standard for accuracy and data quality in the field of immunosequencing.