Fig. 3: Validation of the gMCS pipeline for the calculation of metabolic vulnerabilities in cancer.
From: An automated network-based tool to search for metabolic vulnerabilities in cancer
a Comparison of gene essentiality predictions between tINIT and our gMCSs approach for the 621 cell lines considered in Robinson et al.1 with available CRISPR screening in the DepMap database11. Our gMCSs approach was applied under two gene expression thresholding strategies: gmcsTH5 and localT2. Note: MCC denotes Mathew Correlation Coefficient; ****/***/ns refers to the statistical significance level from an unpaired one-sided Wilcoxon test that compares the gMCS approach against the results obtained with tINIT, respectively ****: p-value ≤ 0.0001, ***: p-value ≤ 0.001, and ns: non-significant. b Computation time of gmctool for calculating essential genes (Single) and lethal DKOs (Double) in the 621 cancer cell lines considering all basic metabolic tasks (All tasks), according to the number of threads used in the processing and gene expression thresholding strategies. Entry times correspond to the mean of 10 computer runs, measured in seconds. Error bands denote the mean ± SEM. Source data are provided as a Source Data file. c Expression heatmap (in TPM) of a relevant gMCS which contains ENO1 and 5 genes: {ENO1, ENO2, ENO3, AMT, PDHA2, ABAT}. Each column corresponds to a single cell line of the 621 considered above.
