Evolution and modulation of antigen-specific T cell responses in melanoma patients

Scripts to reproduce figures and analyses in the manuscript "Evolution and modulation of antigen-specific T cell responses in melanoma patients" Huuhtanen et al., submitted

Installation

Clone this github repository, e.g., in terminal by

$ git clonehttps://github.com/janihuuh/melanomap_manu

All the other dependencies (R-packages, Python modules) are mentioned in the code and their installation guides can be found in their own respective guides. The typical installation time for all the software on a standard laptop (e.g., Apple M1 16Gb) should not exceed 1 hour.

Pseudocode for MelanoMAP

scTCRab-seq preprocessing

scRNAseq preprocessing

for Huuhtanen et al., Durante et al data, do  
  read 10X CellRanger output 
  filter based on the quality of cells 

for Jerby-Arnon et al, Li et al., Sade-Feldman et al, do  
  read count matrix
  filter pre-recieved QC

### No batch correction (Jerby-Arnon et al., Li et al., Sade-Feldman et al)
for scTCRseq data, do  
  read count matrix
  filter ## based on the quality of cells (incomple TCRab information, low confidence data)  
  Normalize and scale 
  Find HVGs ## select top 2000, exclude V(D)J genes
  PCA on HVGs ## retain PCs >2 SD
  Graph-based clustering on PCs
  annotate 

Batch correction (Huuhtanen et al., Durante et al.)

for scTCRseq data, do  
  read count matrix
  filter ## based on the quality of cells (incomple TCRab information, low confidence data)  
  scVI on all cells  ## calculate 30 latent embeddings
  Graph-based clustering on latent embeddings
  annotate 

To reproduce the results:

1) Clone this repository

git clone https://github.com/janihuuh/melanomap_manu
cd path/to/melanomap_manu/

2) Obtain the data


The processed scRNA+TCRab-seq data can be received from EGA (accession number EGAS00001005580) and links listed in Supplementary Data 1.
The TCRb-seq data can be received from immuneAccess from links listed in Supplementary Data 1

3) Create seurat-objects

Rscript R/main.R ## init the helper-functions, coloring, etc.
Rscript R/helper/run_preprocessTCRab.R ## preprocess the TCRab-seq
python python/run_scvi_example.py ## obtain the latent embeddings to use in clustering, UMAPs. This is just an example script
Rscript R/rnaseq/run_createSeurat.R ## read the CellRanger output, filter, merge TCRab-data, merge scVI latent embeddings, run singleR, DEs, pathways, etc.


4) Run additional RNAseq analyses

Rscript R/rnaseq/run_cellphone.R ## init data for CellPhoneDb
bash bash/run_cellphonedb.sh ## Run CellPhoneDb
Rscript R/rnaseq/run_scenic.R ## run Scenic-analysis
bash bash/run_vartrix.sh ## Run Vartrix
Rscript R/rnaseq/run_edgeR.R ## run bulk-RNAseq analysis

5) Run additional TCRseq analyses

bash bash/run_vdjtools.sh ## preprocess the bulk-TCRb-data, subsample, calcualte diverisities, etc.
bash bash/run_gliph2_expample.sh ## run GLIPH2 collectively and on individual samples; this is just an example script
python python/run_tcrgp.py ## run TCRGP collectively on samples
Rscript R/tcrseq/run_antigen_drive.R ## run antigen-drive based analyses