An R package for identifying synthetic lethal pairs from large-scale perturbation screens.
The entire dataset used in the paper is big and cannot be stored on Github. The raw shRNA data has already been published as a part of project DRIVE (https://data.mendeley.com/datasets/y3ds55n88r/4 ) and all the mutation and copynumber data from CCLE are available at https://portals.broadinstitute.org/ccle. The MutSig 2CV v3.1 MAF file for each cancer type is available at http://firebrowse.org/. If you wish to use the processed data, please contact us and we'd be happy to share them. The code for processing the data, running both pan-cancer and cancer-type specific analyses, performing causal inference, and plotting the results are available under Scripts.
You can install SLIdR using devtools
.
install.packages("devtools")
library(devtools)
install_github("cbg-ethz/slidr")
To run SLIdR, specify a path to store the results and use the identifySLHits
function. An example dataset for liver cancer is available in the package under LiverData
.
library(slidr)
library(dplyr)
data(LiverData)
# Path for results
path_results <- "~/Downloads/"
# Threshold for significance in WT cell lines
thresh <- 0.1
hits <- slidr::identifySLHits(canc_data = LiverData,
path_results = path_results,
WT_pval_thresh = thresh)
# Filtering significant hits in WT cell lines
hits <- hits %>%
dplyr::filter(WT_pvalue >= thresh)
The resulting variable hits
is a dataframe of all the SL pairs in liver cancer as reported in the paper. SLIdR creates two folders in the specified output directory:
Hit_List
- including .txt file with all the hits before filtering byWT_pval_thresh
valuePlots
- Boxplots of all the significant SL pairs after filtering byWT_pval_thresh
value
The manuscript is available on bioRxiv. The supplementary files and all the pan-cancer and cancer-type specific hits are available under supplementary material.
Sumana Srivatsa
Hesam Montazeri
If you have any questions, please contact
Sumana Srivatsa