ECLIPSE: Enhanced CLonotypic Inference via Prediction of Single-cell Expression

ECLIPSE is a novel method of scTCR-seq analysis that can predict the clonotypes of T cells possessing >2 or only 1 TCR chain. These populations of cells, which can represent >50% of cells in some samples, are not well suited for traditional scTCR-seq analysis. By mapping chain pairings across the sample and using the EM algorithm, ECLIPSE is able to predict chains lost to sequencing dropout and discern between cells that have >2 TCR due to true
biology vs. sequencing artifacts. Together this leads to larger clone sizes and less cells without a clonotype assignment, allowing users to make best use of their scTCR-seq data.

Authors/Contributors

Ethan C. Burns, Zhaochen Ye, Kelly Street, and David A. Braun

Installation

To install ECLIPSE from Bioconductor, please run:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("ECLIPSE")

To download the development version of ECLIPSE, please run:

remotes::install_github("eburns10/ECLIPSE")

Manuscript

ECLIPSE is described in detail in our pre-print on bioRxiv: "VDJdive and ECLIPSE enhance single-cell TCR sequencing analysis through the probabilistic resolution of ambiguous clonotypes" available here

Issues

Please use our issues page or email ethan.burns [@] yale.edu

How to use

Please view the vignette in the vignettes folder