/depmap-analysis

This repostory contains codes for Shimada et al., "A tool for browsing the Cancer Dependency Map reveals functional connections between genes and helps predict the efficacy and selectivity of candidate cancer drugs"

Primary LanguageR

Analysis of DepMap for generating shinyDepMap

This repostory contains codes to find out 'classes' of essential modalities of genes from the DepMap project data of the Broad Intitute. (Note: this repo is updated in Dec 2020. Now the dependency scores are calculated with various mixing ratios between CRISPR and shRNA. For the details, see an upcoming publicaiton.)

01 pre processing

Download and preprocess associated files from DepMap project.

02 compute perturbation score

Impute and combine CRISPR and shRNA dependency scores. (Updated: now the dependency scores are defined using multple mixing ratios of CRISPR and shRNA)

03 find essential genes

Computed two parameters, 'efficacy' and 'selectivity', and the essential genes from the combined perturbation scores are identified for each mixing ratio.

04 adam

Lineage-dependent and common gene essentiality are computed using ADaM2 algorithm.

05 gsea

Pathways that are overrepresented among generally or selectively essential genes are identified using gene set enrichment analysis.

06 echodots

ECHODOTS, a tSNE and DBSCAN-based cluster analysis, is performed among essential genes. See 06 echodots for the details.

07 prep for shinyDepMap

Files to be loaded in shinyDepMap are computed.