T cell receptor repertoire signatures associated with COVID-19 severity

  1. Jonathan J. Park
  2. Kyoung A V. Lee
  3. Stanley Z. Lam
  4. Sidi Chen

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

T cell receptor (TCR) repertoires are critical for antiviral immunity. Determining the TCR repertoires composition, diversity, and dynamics and how they change during viral infection can inform the molecular specificity of viral infection such as SARS-CoV-2. To determine signatures associated with COVID-19 disease severity, here we performed a large-scale analysis of over 4.7 billion sequences across 2,130 TCR repertoires from COVID-19 patients and healthy donors. TCR repertoire analyses from these data identified and characterized convergent COVID-19 associated CDR3 gene usages, specificity groups, and sequence patterns. T cell clonal expansion was found to be associated with upregulation of T cell effector function, TCR signaling, NF-kB signaling, and Interferon-gamma signaling pathways. Machine learning approaches accurately predicted disease severity for patients based on TCR sequence features, with certain high-power models reaching near-perfect AUROC scores across various predictor permutations. These analyses provided an integrative, systems immunology view of T cell adaptive immune responses to COVID-19.

Article activity feed

  1. PREreview

    This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/5933728.

    Main Claim & Relevance:

    In this preprint by Park et al., a large scale analysis of T cell receptor repertoire signatures was performed in order to link the TCR repertoire signature to COVID-19 infection severity. Antigen exposure from COVID-19 significantly decreased the diversity of repertoires and reshaped clonal representation. Machine learning algorithms were then trained on the data obtained from the TCR repertoires, in order to predict COVID-19 infection severity in patients. These algorithms were able to accurately predict the severity of a COVID case, however they were more effective in predicting mild and moderate disease than severe disease.

     

    Are the findings strong, reliable, potentially informative, not informative, or misleading?

    The findings are reliable. The TCR samples were obtained from several different groups from around the world, and the sample size of 2,130 individuals adds to the strength of the findings. The use of the data obtained through TCR repertoires in machine learning algorithms will need to be confirmed through further studies, as well as through the use of additional algorithms.

     

    How might these ideas presented by the main claims further knowledge of the COVID-19 Pandemic?

    Currently, there is a lack of literature regarding TCR specificity groups for COVID-19, and this paper provides useful information as to the impact of COVID antigens on TCR repertoires. This study claims that TCR repertoire data can be used alongside machine learning algorithms to predict the severity of a COVID-19 case. If this is the case, this could prove to be a powerful prognostic tool and aid in patient care, however these findings will need to be verified through additional studies.

  2. ScreenIT

    SciScore for 10.1101/2021.11.30.470640: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Single cell TCR-seq and gene expression (GEX) data for CD4+ and CD8+ T cell repertoires from COVID-19 patients and healthy donors from the ISB-Swedish COVID-19 Biobanking Unit 25 was obtained from the ArrayExpress database 42 (http://www.ebi.ac.uk/arrayexpress) using the accession number E-MTAB-9357.
    ArrayExpress
    suggested: (ArrayExpress, RRID:SCR_002964)
    Immune repertoire statistics: Clonotype statistics and diversity metrics were calculated using Immunarch v0.6.6 44.
    Immunarch
    suggested: None
    Single cell transcriptome analyses: Single cell transcriptome data from the ISB-S dataset were processed using Seurat v4.0.4.
    Seurat
    suggested: (SEURAT, RRID:SCR_007322)
    Upregulated or downregulated genes with significance q-value < 1e-4 were then used for functional annotation with DAVID analysis.
    DAVID
    suggested: (DAVID, RRID:SCR_001881)
    Training and evaluation of machine learning models: Five ML-based approaches were trained on the k-mer frequency matrix generated from amino acids in the CDR3 region in the T cell repertoires of healthy donor and COVID-19 patients from the ISB-S datasets, using Python v3.8.6 and scikit-learn v0.23.1.
    Python
    suggested: (IPython, RRID:SCR_001658)
    scikit-learn
    suggested: (scikit-learn, RRID:SCR_002577)
    Plotly v5.1.0 was used to generate ROC plots from performance results.
    Plotly
    suggested: (Plotly, RRID:SCR_013991)
    Graphical illustrations: Certain graphical illustrations were made with BioRender (biorender.com).
    BioRender
    suggested: (Biorender, RRID:SCR_018361)

    Results from OddPub: Thank you for sharing your code and data.

    Results from LimitationRecognizer: An explicit section about the limitations of the techniques employed in this study was not found. We encourage authors to address study limitations.

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: Please consider improving the rainbow (“jet”) colormap(s) used on pages 29 and 30. At least one figure is not accessible to readers with colorblindness and/or is not true to the data, i.e. not perceptually uniform.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  3. Version published to 10.1101/2021.11.30.470640v1 on bioRxiv