Biology subjects

Larman, H. B.

Publications and source records attributed to Larman, H. B..

21 records · Page 2Linked to original sources

bioRxiv2018

Ultra-Efficient Short Read Sequencing of Immune Receptor Repertoires

AbstractT cell receptor (TCR) repertoire sequencing is increasingly employed to characterize adaptive immune responses. However, current TCR sequencing methodologies are complex and expensive, limiting the scale of feasible studies. Here we present Framework Region 3 AmplifiKation sequencing (FR3AK-seq), a simplified multiplex PCR-based approach for the ultra-efficient analysis of TCR complementarity determining region 3 (CDR3) repertoires. By using minimal primer sets targeting a conserved region adjacent to CDR3, undistorted amplicons are analyzed via short read, single-end sequencing. We find that FR3AK-seq is sensitive and quantitative, performing comparably to two industry standards. FR3AK-seq was utilized to quickly and inexpensively characterize the T cell infiltrates of muscle biopsies obtained from 145 patients with idiopathic inflammatory myopathies and controls. A cluster of related TCRs was identified in samples from patients with sporadic inclusion body myositis, suggesting the presence of a shared antigen-driven response. The ease and minimal cost of FR3AK-seq removes critical barriers to routine, large-scale TCR CDR3 repertoire analyses.

immunology↗

bioRxiv2018

DeepTCR: a deep learning framework for revealing structural concepts within TCR Repertoire

Deep learning algorithms have been utilized to achieve enhanced performance in pattern-recognition tasks, such as in image and vocal recognition 1,2. The ability to learn complex patterns in data has tremendous implications in the genomics and immunology worlds, where sequence motifs become learned features that can be used to predict functionality, guiding our understanding of disease and basic biology 3-6. T-cell receptor (TCR) sequencing assesses the diversity of the adaptive immune system, where complex structural patterns in the TCR can be used to model its antigenic interaction. We present DeepTCR, a broad collection of unsupervised and supervised deep learning methods able to uncover structure in highly complex and large TCR sequencing data by learning a joint representation of a given TCR by its CDR3 sequences, V/D/J gene usage, and HLA background in which the T-cells reside. We demonstrate the utility of deep learning to provide an improved featurization of the TCR across multiple human and murine datasets, including improved classification of antigen-specific TCRs in both unsupervised and supervised learning tasks, understanding immunotherapy-related shaping of repertoire in the murine setting, and predicting response to checkpoint blockade immunotherapy from pre-treatment tumor biopsies in a clinical trial of melanoma. Our results show the flexibility and capacity for deep neural networks to handle the complexity of high-dimensional TCR genomic data for both descriptive and predictive purposes across basic science and clinical research.

bioinformatics↗

bioRxiv2018

Improved Analysis of Phage ImmunoPrecipitation Sequencing (PhIP-Seq) Data Using a Z-score Algorithm

Phage ImmunoPrecipitation Sequencing (PhIP-Seq) is a massively multiplexed, phage-display based methodology for analyzing antibody binding specificities, with several advantages over existing techniques, including the uniformity and completeness of proteomic libraries, as well as high sample throughput and low cost. Data generated by the PhIP-Seq assay are unique in many ways. The only published analytical approach for these data suffers from important limitations. Here, we propose a new statistical framework with several improvements. Using a set of replicate mock immunoprecipitations (negative controls lacking antibody input) to generate background binding distributions, we establish a statistical model to quantify antibody-dependent changes in phage clone abundance. Our approach incorporates robust regression of experimental samples against the mock IPs as a means to calculate the expected phage clone abundance, and provides a generalized model for calculating each clones expected abundance-associated standard deviation. In terms of bias removal and detection sensitivity, we demonstrate that this z-score algorithm outperforms the previous approach. Further, in a large cohort of autoantibody-defined Sjogrens Syndrome (SS) patient sera, PhIP-Seq robustly identified Ro52, Ro60, and SSB/La as known autoantigens associated with SS. In an effort to identify novel SS-specific binding specificities, SS z-scores were compared with z-scores obtained by screening Ropositive sera from patients with systemic lupus erythematosus (SLE). This analysis did not yield any commonly targeted SS-specific autoantigens, suggesting that if they exist at all, their epitopes are likely to be discontinuous or post-translationally modified. In summary, we have developed an improved algorithm for PhIP-Seq data analysis, which was validated using a large set of sera with clinically characterized autoantibodies. This z-score approach will substantially improve the ability of PhIP-Seq to detect and interpret antibody binding specificities. The associated Python code is freely available for download here: https://github.com/LarmanLab/PhIP-Seq-Analyzer.

bioinformatics↗