nt12dreuar3esd

2

nature research | reporting summary

April 2018

Data

Policy information about availability of data

All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:

• Accession codes, unique identifiers, or web links for publicly available datasets


• A list of figures that have associated raw data


• A description of any restrictions on data availability
  FASTQ files containing raw reads from the scRNA-seq and scTCR-seq analyses have been deposited with the European Genome-phenome Archive (GA) under
  studies EGAS00001003993 and EGAS00001003994, and datasets EGAD00001005464 and EGAD00001005465. These files are available under controlled access
  upon request to the Data Access Committee, with contact information provided at the corresponding Web page at EGA. Processed output files from Cell Ranger,
  integrated assay results from Seurat, and metadata with UMAP coordinates, cluster assignments, and clonotypes are available from the NCBI Gene Expression
  Omnibus database under accession GSE139555, which also provides computer code used to generate the analyses and figures in this paper as a Supplementary File.

Field-specific reporting

Please select the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences

For a reference copy of the document with all sections, see nature.com/authors/policies/ReportingSummary-flat.pdf

Life sciences study design

All studies must disclose on these points even when the disclosure is negative.

Sample size Sample size of patients was determined in part by the exploratory nature of our study, the novelty of the 10x Genomics single-cell TCR-seq

assay technology, and the relative difficulty of obtaining surgical samples. However, our study of 14 patients is comparable to published

studies of single-cell RNA-seq and single-cell TCR-seq data, with 2018 studies having 12–14 patients using another technology (SmartSeq2).

Furthermore, our dataset has greater sequencing depth than published studies, providing evidence that our sample sizes of cells and clones

should be adequate for comparable analysis of T cell subtypes and clonal expansion. Sample sizes for clinical trial analyses were determined

by the designers of the clinical trials.

Data exclusions Single cells from CD3- and CD45-selected samples were separated computationally into T and non-T cells, as described in Methods and

documented in Extended Data Fig 1a–g. Patient bcc.su003 from Yost et al., 2019 showed no overlapping clones between pre- and post-

treatment tumour, and could not be analyzed further, as described in Methods.

Replication We tested our script for identifying clonotypes against the assignment from CellRanger software for each patient. We compared our cluster

analysis of T cell subsets against the clusters obtained by Guo et al., 2018, Zhang et al., 2018, and Yost et al., 2019. We confirmed our

relationship between peripheral clonal expansion and infiltration, as well as the correlation between blood clone size and tumor clone size,

using 14 patients from Guo et al., 2018 and 12 patients from Zhang et al., 2018. We observed relationships between T cell subsets and clonal

expansion patterns similar to ours using the same two datasets. We observed a correlation between blood transcript counts and novel

tumour clone size on two patients in Yost et al., 2019 with bulk RNA-seq measurements in blood and the largest numbers of novel clones.

We confirmed our hypotheses regarding the replenishment of T cells from blood using 6 patients from Yost et al., 2019. We confirmed our

analysis of viral reactivity using two datasets: VDJdb and TCGA-derived TCR sequences. We performed survival analysis on gene signatures for

tissue expansion patterns using three independent clinical trials.

Randomization Single-cell TCR data were grouped using T cell receptor chains and CDR3 nucleotide sequences assigned by the Cell Ranger software package.

Single-cell RNA-Seq data were grouped using existing classification algorithms from the Seurat package. All two-dimensional gene expression

maps plotted cells in random order. Random jitter was added to scatterplots to visually display all points.

Blinding Patient samples were obtained from commercial vendors, who were blinded to the scientific purpose of the samples. Sample preparation

was performed by laboratory personnel who were blinded to the subsequent computational analysis. Exploratory computational analyses

were performed in an unsupervised manner, with analyses blinded to preconceived notions of the expected outcome. Statistical tests were

performed without bias about expected results. Clinical data were obtained from randomized clinical trials that blinded both patients and

practitioners about their treatment arms.

Reporting for specific materials, systems and methods