Code repository for preprint 'Dimensionality reduction methods for extracting functional networks from large-scale CRISPR screens'
DOI: https://doi.org/10.1101/2023.02.22.529573
Repository for autoencoder-normalization code: https://github.com/csbio/ae-norm
Repository for input data and output files : https://zenodo.org/record/7671685#.Y_gi9nbMK5c
-Download ONION-git repo.
-All source codes (R scripts) are in the ONION-git/src directory.
-Create 'data' directory in ONION-git.
-Input data required for this are uploaded here - https://zenodo.org/record/7671685#.Y_gi9nbMK5c
-Create 'output' directory in ONION-git. Create 'ae','pca','rpca' and 'onion' directories in the 'output' directory.
-Download FLEX and put in the same directory as the ONION-git.
-Required directory and file organization
.
├── ONION-git
│ ├── src
│ │ ├── ... # provided R scripts
│ ├── data
│ │ ├── 20Q2_GLS_fdr.npy
│ │ ├── 20Q2_GLS_sign.npy
│ │ ├── Achilles_gene_effect.csv
│ │ ├── depmap_q2_2020_nona_mean.tsv
│ │ ├── depmap_q2_2020_nona_mean_rst_clp_mms.tsv
│ │ ├── all_genes_20q2.txt
│ │ ├── Mitochondial_genelist_1_26_2021_genes.tsv
│ │ ├── olfactory_receptors.csv
│ │ ├── ae_tanh_e1_depmap_20q2
│ │ │ ├── 20q2_epochs_1_latent_1_normalized_ae.tsv
│ │ │ ├── 20q2_epochs_1_latent_2_normalized_ae.tsv
│ │ │ ├── 20q2_epochs_1_latent_3_normalized_ae.tsv
│ │ │ ├── 20q2_epochs_1_latent_4_normalized_ae.tsv
│ │ │ ├── 20q2_epochs_1_latent_5_normalized_ae.tsv
│ │ │ ├── 20q2_epochs_1_latent_10_normalized_ae.tsv
│ ├── output
│ │ ├── ae
│ │ │ ├── ... # outputs are created after running R scripts
│ │ ├── pca
│ │ │ ├── ... # outputs are created after running R scripts
│ │ ├── rpca
│ │ │ ├── ... # outputs are created after running R scripts
│ │ ├── onion
│ │ │ ├── ... # outputs are created after running R scripts
├── FLEX
│ ├── ... # Directories from FLEX package
-Required packages and dependencies should be installed prior to running the scripts. Required packages are listed in the 'R packages'.
-Run from inside src directory in the given order. Some scripts are dependent on outputs from other scripts. Please refer to script_dependency_flow_.pdf for script dependency.
-Please refer to data_flow_.pdf and data_flow_2_.pdf for script input/output flow.
Rscript pre_process.R
Rscript pca_normalization_pipeline.R
Rscript rpca_normalization_pipeline.R
Rscript ae_normalization_pipeline.R
Rscript network_correlation_pipeline.R
Rscript pca_onion_pipeline.R
Rscript rpca_onion_pipeline.R
Rscript ae_onion_pipeline.R
Rscript pre_process_gls.R
Rscript onion_evaluation_pipeline.R
Rscript AUPRC_barplot_layer_evaluation_pipeline.R
Rscript AUPRC_barplot_onion_evaluation_pipeline.R
Pre-process DepMap data:
pre_process.R
PCA-normalization:
pca_normalization_pipeline.R
RPCA-normalization:
rpca_normalization_pipeline.R
AE-normalization:
ae_normalization_pipeline.R
PC-Onion:
pca_onion_pipeline.R
RPC-Onion:
rpca_onion_pipeline.R
AE-Onion:
ae_onion_pipeline.R
Evaluation:
network_correlation_pipeline.R
pre_process_gls.R
onion_evaluation_pipeline.R
AUPRC_barplot_layer_evaluation_pipeline.R
AUPRC_barplot_onion_evaluation_pipeline.R