sumo_analysis

Analysis and figures for SUMO manuscript.

Set-up instructions

1. Clone this repository and its submodule

git clone --recurse-submodules https://github.com/ratan-lab/sumo_analysis.git

2. Install SUMO from command line (please note that the package require python3.6+):

python3 -m pip install --upgrade pip
python3 -m pip install python-sumo

3. Install packages from R console:

install.packages(c('PMA', 'PINSPlus', 'R.matlab', 'devtools', 'Matrix', 'rticulate', 'cluster', 'survival', 
                   'tidyverse', 'ggsci', 'ggpubr', 'cowplot', 'gridExtra'))

library("devtools")
install_github("danro9685/CIMLR", ref = 'R')
install_github("saezlab/progeny")

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install(c("iClusterPlus", "DESeq2"))

4. Download and install source packages of SNFtool and LRAcluster:

install.packages('LRAcluster_1.0.tgz', repos = NULL, type="source")
install.packages('SNFtool_2.0.3.tar.gz', repos = NULL, type="source")

Comparing tool performance on various cancer datasets

All the scripts should be run from the sumo_analysis/benchmark directory.

1. Download benchmark data from http://acgt.cs.tau.ac.il/multi_omic_benchmark/download.html.
2. Extract all .zip files into the sumo_analysis/benchmark/data directory.
3. Run run_benchmark.R script to compare tool performance.
4. Run test_pathways.R to compare pathway activity between found clusters (script generates benchmark_pathway_activity.tsv).
5. To create Fig2 run plot_benchmark.R script.
6. Run create_supp_table.R script to create summary table (Supplementary Table S2).

Stability of benchmark results

1. From sumo_analysis/benchmark directory run run_benchmark_eval.R script.
2. To create FigS4 run plot_benchmark_eval.R script.

Comparing tool performance on simulated datasets

All the scripts should be run from the sumo_analysis/simulations directory.

1. Run run_noisy_simulation.R script to compare tool performance on datasets with various level of noise.
2. Run run_missing_simulation.R script to compare tool performance on incomplete datasets (containing missing data).
3. To create Fig1 and FigS2 run plot_simulations.R script.