Analysis code for the MetTag lineage-barcoding study of triple-negative breast cancer (TNBC) lung metastasis, accompanying Aleksandrovic et al., bioRxiv 2025 (doi: 10.1101/2025.08.13.669778). This citation will be updated to the peer-reviewed Nature Communications reference upon publication.
This repository contains the analysis code for two single-cell RNA-seq experiments using cell hashing (HTO) and custom feature-barcode (LARRY-style) clonal lineage tracing.
R16 (October 2021) Mixed barcode/cancer-cell-species experiment to test barcode detection at the single-cell level and evenness of distribution both in vivo and in vitro. Four samples/HTOs were pooled: two lung-metastasis samples sorted for GFP+ cells (each injected with a mixture of all six BC.IDs), one E0771 cell-culture pool of all BC.IDs, and one MDA-MB-231 human cancer cell line expressing a single BC.ID for QC. Feature-barcode libraries were amplified using the 10x Genomics 3' Chromium feature-barcoding protocol. This QC dataset informed the choice of the three most even BC.IDs used in subsequent lung metastasis experiments.
R18 (March 2022) CITE-seq (10x Genomics) profiling of transcriptomes and clonality of E0771 mouse cancer cells injected sequentially into the retro-orbital sinus of wild-type C57BL/6 mice and harvested from lungs. Three samples/M-HTOs were pooled onto the 10x Chromium instrument, with three separate libraries per lane (HTO, cell barcode/LARRY, transcriptome). BC.ID1, BC.ID2, and BC.ID6 mark early, mid, and late seeding timepoints, respectively. This sequential-seeding pilot used two ChipG lanes to maximize recovery of rare clones (target: 20,000 cells) and a custom nested-PCR method to amplify the feature-barcode library.
| Script | Description |
|---|---|
Code/01_Generate_CellBC.R |
HTO demultiplexing and QC filtering to generate the singlet cell-barcode whitelist (R18) |
Code/02_Basic_Seurat.R |
Baseline Seurat processing: normalization, PCA, UMAP, clustering, cluster marker identification (R18) |
Code/03_Assign_LARRY.R |
LARRY feature-barcode assignment and adaptive-threshold singlet/doublet/multiplet classification (R18) |
Code/04_General_BC_analysis_EA.R |
Immune vs. cancer cell gating, subclustering, and differential expression analysis (R18) |
Code/05_GSVA_analysis_EA.R |
ssGSEA/GSVA pathway-level analysis of cancer cell subsets (R18) |
Code/06_R16_BC_transcriptome_preprocessing.R |
HTO demultiplexing, QC, and LARRY/Lib_ID barcode assignment (R16) |
Code/07_R16_BCID_transcriptome_analysis.R |
In vivo vs. in vitro differential expression and pathway enrichment analysis (R16) |
Scripts are numbered in the order they were run. Each script begins with a base_dir/setwd() line marked <-- EDIT that must be set to the local directory containing the downloaded data before running.
Raw sequencing data for both experiments (R16, R18) are deposited in NCBI GEO under accession GSE309120.
Analyses were performed in R using the following packages: Seurat, dplyr, ggplot2, sctransform, infercnv, patchwork, viridis, Matrix, RColorBrewer, ggrepel, scales, plotly, GSEABase, GSVA, GSVAdata, Biobase, limma, parallel, readxl, tidyr, pheatmap, ggpubr, clusterProfiler, org.Mm.eg.db, ReactomePA, msigdbr.
If you use this code, please cite:
Aleksandrovic E, Fross SR, Golomb SM, Liu X, Zhao Z, Das NM, Reese TC, Ma W, Lopez J, Stack MS, Zhao M, Zhang S. Temporal Clonal Tracing and Functional Perturbation Reveal Niche-Adaptive and Tumor-Intrinsic IFNγ Dependencies Driving Ovarian Cancer Metastasis. bioRxiv. 2025. doi: 10.1101/2025.08.13.669778
This is a preprint citation; it will be replaced with the formal Nature Communications citation upon publication.
Released under the MIT License.
Siyuan Zhang, UT Southwestern Medical Center (Siyuan.Zhang@UTSouthwestern.edu)