Thesis-Project: Identifying metabolic shifts in inflamed endothelial cells using Genome-Scale Metabolic model approaches

This repository contains code for a Master’s thesis investigating metabolic changes in inflamed endothelial cells using genome-scale metabolic modeling. The analysis combines Cellfie and context-specific modeling applied to RNA-seq datasets.

How to run

Requirements

• MATLAB (2020a or newer)
• COBRA Toolbox
• Human-GEM models (1 and 2)
• MetabolicTasksDCM (Cellfie)
• • RNA-seq datasets from Gene Expression Omnibus (GEO):
  • GSE118815
  • GSE131590
  • GSE184512

Setup

1. Clone this repository
2. Add all folders in /Matlab/functions to the MATLAB path
3. Install and initialize the COBRA Toolbox

Workflow Cellfie

1. Run preparotary steps Cellfie (see /Matlab/scripts/Cellfie/task_analysis.m) to generate taskReport and essential reaction list
2. Run Cellfie analysis (/Matlab/scripts/Cellfie/task_analysis_on_data.m) to generate Cellfie scores
3. Analyse, visulaise and statistical testing using R (R/cellfieScoresVisualizations.Rmd)

Workflow context-specific models

1. Context-specific models are created using CORDA2 (see Matlab/scripts/generateGEMsCORDA.mlx)
2. Results are stored in /Results/reconstructions
3. Reaction content are analysed with analyseCORDAmodels.mlx
4. FBA experiments data is generated with FBAExperiments.mlx
5. Visualisation of FBA was done in R using plotFluxesFBA.Rmd

Data folder structure

• raw_counts: raw RNA-seq counts
• processed: cleaned and normalized data
• TPM/soft: transcript-per-million normalized data
• metadata: sample annotations
• Mediums: media conditions for metabolic models
• Task_analysis: Cellfie task list