This repo contains the code used to analyze the data from Alvarez, Benhammou et al. 2021.

The file run_all.sh gives the order in which to run the scripts. The scripts to run the analysis are found in the bin directory.

To download the data from Aizarani et al. and Sharma et al., run the scripts in the data-raw directory.

For the NAFLD-related HCC set, the scripts use the raw counts from STARSolo. The DIEM-filtered counts can be downloaded from GEO under accession number GSE189175. The DIEM filtering scripts can be skipped and the code for clustering can read in the matrices downloaded from GEO.

Although sequencing data is not available for the NAFLD-related HCC data set, the alignment scripts are provided in bin/align.

Set up the conda environment using the bin/create_conda_envs.sh and bin/add_conda_pkg.sh scripts.

The scripts in bin/aizarani2019, bin/sharma2019, and bin/d_sct_cca_all run clustering on the Aizarani et al healthy liver scRNA-seq reference data, the Sharma et al HCC data, and the Rao et al NAFLD-related HCC data, respectively.

The scripts in bin/sharma_aiz run the integrated clustering across the three cohorts. The scripts in bin/tcga_hcc run the TCGA decomposition analysis, and finally the bin/GSE14520 run the LCI decomposition analysis.

To reproduce the manuscript figures and tables, run the scripts in the bin/manuscript folder.