Data used in SeuratIntegrate paper

This repository gathers the data and code used to generate hepatocellular carcinoma analyses in the paper presenting SeuratIntegrate. It contains the scripts to reproduce the figures presented in the article. Some figures are also available as pdf files.To be able to fully reproduce the results from the paper, one shoud: download all the files install R 4.3.3, with correspondig base R packages (stats, graphics, grDevices, utils, datasets, methods and base) install R packages listed in the file sessionInfo.txt install the provided version of SeuratIntegrate. In an R session, run: remotes::install_local("path/to/SeuratIntegrate_0.4.1.tar.gz") install (mini)conda if necessary (we used miniconda version 23.11.0) install the conda environments (if it fails with the *package-list.yml files, use the *package-list-from-history.yml files instead): conda env create --file SeuratIntegrate_bbknn_package-list.yml conda env create --file SeuratIntegrate_scanorama_package-list.ymlconda env create --file SeuratIntegrate_scvi-tools_package-list.ymlconda env create --file SeuratIntegrate_trvae_package-list.yml open an R session to make the conda environments usable by SeuratIntegrate: library(SeuratIntegrate) UpdateEnvCache("bbknn", conda.env = "SeuratIntegrate_bbknn", conda.env.is.path = FALSE) UpdateEnvCache("scanorama", conda.env = "SeuratIntegrate_scanorama", conda.env.is.path = FALSE)UpdateEnvCache("scvi", conda.env = "SeuratIntegrate_scvi-tools", conda.env.is.path = FALSE)UpdateEnvCache("trvae", conda.env = "SeuratIntegrate_trvae", conda.env.is.path = FALSE) Once done, running the code in integrate.R should produce reproducible results. Note that lines 3 to 6 from integrate.R should be adapted to the user's setup.integrate.R is subdivided into six main parts: Preparation: lines 1-56 Preprocessing: lines 58-74 Integration: lines 76-121 Processing of integration outputs: lines 126-267 Scoring of integration outputs: lines 269-353 Plotting: lines 380-507 Intermediate SeuratObjects have been saved between steps 3 and 4 and 5 and 6 (liver10k_integrated_object.RDS and liver10k_integrated_scored_object.RDS respectively). It is possible to start with these intermediate SeuratObjects to avoid the preceding steps, given that the Preparation step is always run before.