covid_metabolism

This repository contains the code for performing the analyses in:

Cheng K, Martin-Sancho L, Pal LR, et al. Genome-scale metabolic modeling reveals SARS-CoV-2-induced metabolic changes and antiviral targets. Mol Syst Biol. 2021;17(11):e10260.

1. Prerequisites

Softwares

• R (version 3.6.3 was used in our study)
• R packages
  • ImNotaGit/my.utils
  • ruppinlab/Rcplex2: required to run genome-scale metabolic modeling (GEM); otherwise may be omitted
  • ruppinlab/gembox
  • other needed packages can be obtained from CRAN or Bioconductor, please see the R scripts
• IBM ILOG CPLEX Optimization Studio or Gurobi: required to run GEM; otherwise may be omitted (CPLEX 12.10 was used in our study)

Data

The required data can be downloaded by running the dnload.sh bash script in the data folder. dnload.sh uses wget and gdown (the latter used for downloading files from Google Drive and can be installed with pip install gdown). Data files containing results from the various analyses are also included in the download.

Alternatively, one can manually download the data files from this Google Drive link then decompress them into the data folder. The two single-cell RNA-sequencing datasets from Liao et al. and Chua et al. are quite large and not included in this download. Check dnload.sh for details on how these two datasets should be downloaded.

2. Description of folders and files

The scripts should be run in the same order as they are introduced below. Each script should be run from the directory it is located in as the working directory.

data

All required data should be saved in this folder.

• dnload.sh: a bash script for downloading the required data
• collect.validation.data.R: prepare the data for validating the MTA prediction

expression

Scripts for gene expression-level analysis:

• de.and.gsea.R: differential expression (DE) and gene set enrichment analysis (GSEA) between the SARS-CoV-2-infected samples and non-infected controls in each dataset
• de.and.gsea.new.R: DE and GSEA between the SARS-CoV-2-infected samples and non-infected controls in each dataset -- modified such that DESeq2 was consistently used for most of the bulk RNA-seq datasets
• de.and.gsea.remdesivir.R: DE and GSEA analysis for the Vero E6 cell remdesivir treatment data

GEM

Scripts for genome-scale metabolic modeling (GEM) analysis with Recon 1:

• prepare.data.R: prepare data for GEM
• imat.and.mta.R: run iMAT and rMTA on each dataset
• dflux.R: differential flux analysis between the SARS-CoV-2-infected samples and non-infected controls in each dataset
• collect.results.R: collect the differential flux analysis and rMTA results across datasets

GEM_Recon3D

Scripts for GEM analysis with Recon 3D:

• prepare.data.R: prepare data for GEM
• imat.and.mta.R: run iMAT and rMTA on each dataset
• dflux.R: differential flux analysis between the SARS-CoV-2-infected samples and non-infected controls in each dataset
• collect.results.R: collect the differential flux analysis and rMTA results across datasets

GEM_remdesivir

Scripts for GEM analysis on the Vero E6 cell remdesivir treatment data with Recon 1:

• prepare.data.R: prepare data for GEM
• imat.R: run iMAT for each experimental group
• mta.R: run rMTA for pairs of experimental groups
• dflux.R: differential flux analysis between pairs of experimental groups
• collect.results.R: collect the differential flux analysis results

GEM_remdesivir_Recon3D

Scripts for GEM analysis on the Vero E6 cell remdesivir treatment data with Recon 3D:

• prepare.data.R: prepare data for GEM
• imat.R: run iMAT for each experimental group
• mta.R: run rMTA for pairs of experimental groups
• dflux.R: differential flux analysis between pairs of experimental groups
• collect.results.R: collect the differential flux analysis results

analyze_results

Scripts and R notebooks for inspecting the results from various analyses and for validating the rMTA predictions:

• functions.R: various functions, will be sourced in check.mta.results.R and check.mta.results.remdesivir.R
• check.mta.results.R: various validations and pathway enrichment analysis of rMTA results
• check.mta.results.remdesivir.R: various validations and pathway enrichment analysis of rMTA results for predicting combinatory target with remdesivir
• Various .Rmd files: R notebook for inspecting and visualizing results from different analyses, see their contents for details