| editor_options | ||||
|---|---|---|---|---|
|
This repository is required to download and preprocess the PIC-seq data from the paper referenced below. The code currently contained will only help in analyses related to the human PIC-seq data available at GSE160903.
- scripts/get_geo_files.R
The purpose of this file is to download the the GEO files and prepare
them for preprocessing, one issue however is you need to copy and paste
the shell commands into the terminal, I was unable to use system() to
get them working properly.
- scripts/human_data_selector.R
This script processes the data as in the paper, creates several outputs files, and sends them to the cloud.
Cloud location:
- output/immunitas_output/singlet_myeloid_pic.mtx: A sparse matrix of the umi counts for singlets sorted on myeloid markers
- output/immunitas_output/singlet_cd3_pic.mtx: A sparse matrix of the umi counts for singlets sorted on CD3+ cells
- output/immunitas_output/doublet_cd3_myeloid_pic.mtx: A sparse matrix of the umi counts for doublets sorted on CD3+ and myeloid markers
- output/immunitas_output/metacell_annos_singlets.tsv: A tsv with annotations of which metacells do singlets fall into (and their cell type)
- output/immunitas_output/parsed_pics_cell_types.tsv: A tsv labeling the PIC cells what their most likely myeloid and t-cell type are.
- output/immunitas_output/metacell_singlet_average_umi_exprs.tsv: A tsv file of the average umi expression across singlets contained in a metacell. From the sin_cl object in the @e_gc slot.
- output/immunitas_output/metacell_singlet_proportion_umi_exprs.tsv: A tsv file of the proportional umi expression across singlets contained in a metacell. From the sin_cl object in the @e_cov slot.
- output/immuntias_output/metacell_annos_metacells.tsv: A tsv labeling the singlet metacells with their respective cell types.
This repository contains the PIC-seq all the needed code and metadata files to analyze and generate figures for Cohen M. & Giladi A. et al. Nature Cancer 2022
In order to run the scripts, download processed data from GSE160903 to the folder output/umi.tab
Unzip and change file names by running these shell commands (in output/umi.tab): gzip -d * ls -1 | awk -F'_' '{print $0,$2}' | xargs -n 2 mv
Download published data into their respective folders: GSE123139 output/published_data/melanoma/ EGAD00001006608 output/published_data/breast/ GSE135382 output/published_data/pic-seq/
To start analysis, run from the root directory: Rscript scripts/run.r
Please send questions to Amir Giladi: aygoldberg@gmail.com
Required R packages:
Package version glmnet 2.0-16 foreach 1.4.4 Matrix 1.2-18 compositions 1.40-2 bayesm 3.1-0.1 energy 1.7-5 robustbase 0.93-3 tensorA 0.36.1 gplots 3.0.1.1 plotrix 3.7-4 plyr 1.8.4 RANN 2.6.1 reshape2 1.4.3 KernSmooth 2.23-15 dendextend 1.9.0 Hmisc 4.2-0 Formula 1.2-3 survival 3.2-3 lattice 0.20-38 tglkmeans 0.2.0 ggrepel 0.8.1 ggplot2 3.3.2.9000 ape 5.2 scales 1.0.0 metacell 0.3.41 tgstat 2.3.5 misha 4.0.10