Single Cell RNAseq BAM files and metadata "Tailoring vascular phenotype through AAV T therapy promotes anti-tumor immunity in glioma, Ramachandran et al, 2023"

Published: 17 April 2023| Version 2 | DOI: 10.17632/fwczkb6xw3.2


This study was designed to capture the changes in CD8 T cell phenotypes in murine glioma (CT-2A) model post immunotherapy with AAV-LIGHT (TNFSF14). Tumour infiltrating CD45 cells were isolated by flow sorting and subject to targeted single-cell transcriptome sequencing and downstream analysis. 1. The 3 .bam files contain aligned sequence data for sc-RNAseq of all the CD8+T cells in the 3 libraries that were analysed in the mentioned study. 2. md5 checksum file to verify integrity of the data 3. Metadata including cell id to annotate the cells to treatment and outcome 4. Normalized counts 4. Normalized gene expression data of CD8 T cells.


Steps to reproduce

PROTOCOLS: Myelin-depleted single-cell suspensions obtained from the brains of individual tumor-bearing mice were sequentially labeled using the mouse Single Cell Sample Multiplexing Kit (Sample-tag) (BD Biosciences, #633793), BD AbSeq Ab-Oligos reagents (Table S6), the sorting antibody CD45-APC (clone I3/2.3) and the viability stain Zombie Aqua (Biolegend, #423101) following the manufacturer’s protocol (BD Biosciences). The stained cell suspension was enriched for live CD45+ immune cells using a BD AriaIII flow sorter (BD Biosciences). Sorted live CD45+ cells from different treatment groups (1x AAV-GFP and 2x AAV-LIGHT per library preparation) were counted and pooled at equal ratios to achieve a total of approximately 20,000 cells and loaded on to the BD Rhapsody cartidge for single cell capoture (BD Biosciences, # 633733, #633731). Tragetted mRNA libraries were generated using the BD Rhapsody Immune Response Panel Mm (BD Biosciences, #633753, # 633774) supplemented with additional genes according to the manufacturers instructions. The quality of the final libraries was assessed using an Agilent 2200 TapeStation with HS D5000 ScreenTape and concentrations were measured by a Qubit Fluorometer . This process was repeated an additional two times to obtain a total of three libraries, each containing samples from 1x AAV-GFP and 2x AAV-LIGHT-treated mice. The three final libraries were diluted to 2nM and pooled for paired-end (Read1: 64bp, Read2: 42bp + i7 Indexes: 8bp) sequencing on NovaSeq 6000 S Prime sequencer. DATA PROCESSING PIPELINE: Data were normalized as counts per 10,000, added an offset of 1 count, and performed Log2 transformation as recommended by the manufacturer in Partek Flow. CD8 T cells were filtered out from other immune cells using CD8a gene expression. Dimensionality reduction using PCA followed by UMAP (default settings) was performed on all genes and proteins. Unsupervised clustering was performed using the Louvain clustering algorithm (using default settings, nearest neighbor type: K-NN, number of nearest neighbors: 30) and biomarkers for each cluster were computed using ANOVA tests with Benjamini-Hochberg test to correct for false discovery rate (FDR) value ≤ 0.05 and log2 fold change ≥1.25. Trajectory analysis (monocle 2), Correlation analysis and AUCell analysis wre all performed in Partek flow using default settings.


Uppsala Universitet


Single-Cell RNA Sequencing, Meta Dataset



2016-02495, 2020-02563, 2019-01326



Knut och Alice Wallenbergs Stiftelse

KAW 2019.0088


PR2018-0148, PR2020-0167, PR2021-0122, TJ 2019-0014


CAN 2017/502, 20 1008 PjF, 20 1010 UsF, 190184Pj