/SL_analysis

Primary LanguageJupyter NotebookGNU General Public License v3.0GPL-3.0

A pan-cancer atlas of tumour development of synthetic lethal gene developmental trajectory changes in tumour-infiltrating bone marrow cells

This repository contains the scripts and function to reproduce the results of SL analysis

Content

  • SL_Function.py: the analysis function of SL
  • data: the SL pair
  • script: the pipeline of SL analysis in pan-cancer
  • crlm: the pipeline of SL cell-cell interaction analysis in calm

Tutorial

A simple tutorial on SL analysis will be given here

from SL_Function import SL_Analysis
import anndata
#import data
adata=anndata.read_h5ad("data/lym_dpt.h5ad")
SL_pd=pd.read_csv('data/sl-for-scrnaseq-full.csv')
#Analysis Cancer 
adata_c=adata[adata.obs['tissue']=='T']
SL_c=SL_Analysis(adata_c,SL_pd)
sl_c.Lazy_analysis(['SL_count.csv'.format(cancer,trajectory),'SL_slope.csv'.format(cancer,trajectory),'SL_pair.csv'.format(cancer,trajectory)])
#Analysis Paracancer
adata_p=adata[adata.obs['tissue']=='N']
SL_p=SL_Analysis(adata_p,SL_pd)
sl_p.Lazy_analysis(['SL_count.csv'.format(cancer,trajectory),'SL_slope.csv'.format(cancer,trajectory),'SL_pair.csv'.format(cancer,trajectory)])

Data

The parsed data can be downloaded in each folder.

Data source

Previously published scRNA-seq data reanalyzed here are available under accession codes GSE138709 (Intrahepatic cholangiocarcinoma (ICC)), PRJNA768891 (clear cell renal cell carcinoma (ccRCC)), E-MTAB-8107 (Ovary/Breast/Colorectal cancer (OC/BRCA/CRC)). The gastric cancer (GC) data were downloaded from http://dna-discovery.stanford.edu/download/1401/. We downloaded scRNA-seq of colorectal cancer liver metastases from GEO (GSE164522).

Contact