Single cell profiling of breast cancer T cells reveals a tissue-resident memory subset associated with improved prognosis



Hover over data points for cluster assignment and to display menu in top right of plot with additional controls, including snapshot download.





Hover over data points for cluster assignment and to display menu in top right of plot with additional controls, including snapshot download. Colour of points represents SAVER imputed expression.

Data description

Primary tumours were resected from two triple-negative breast cancer patients. Dissociated cells were FACS sorted to produce a viable population of CD3+ TILs. Single-cell RNA-seq libiraries were prepared with 10X Genomics Chromium Single Cell 3′ Solution and sequenced on HiSeq 2500 High Output Mode using V4 clustering and sequencing chemistry. Raw data were pre-processed with 10X Genomics CellRanger Software to obtain UMI counts. After quality control, 5,759 (4,844+915) cells and 15,621 genes were retained for downstream analysis.


Data processing

Unimputed data is normalized for library size with Seurat default method. Imputed data is normalized within SAVER. For feature t-SNE plots, data is log-transformed with offset 0.1. For violin plots, data is log-transformed with offset 1.


Data availability

Associated raw and processed data are freely available. Please see the related publication:

https://doi.org/10.1038/s41591-018-0078-7