Investigating metabolic re-wiring in the children cohort using the Genome-Scale Metabolic (GEM) models by incorporating transcriptomics data.

Hello Everyone!! I am sharing some quick insights into this repo. This is my Masters Thesis Project, titled "Investigating metabolic re-wiring in the children cohort using the Genome-Scale Metabolic (GEM) models."

The cellular metabolism of the children cohort was studied using metabolic modelling approaches like Genome-scale metabolic modelling (GEM) and Flux Balance Analysis. Personalized GEMs were created, compared, and grouped based on their similarities. Data-driven patient grouping allows healthcare providers to analyze large amounts of patient data to identify patterns, use predictive analytics, and anticipate potential health risks. This personalized approach improves patient outcomes by ensuring that treatments are more effective and targeted. It plays a vital role in population health management.

We have used transcriptomics data from 150 children living in different socio-economic environments in South Africa. Context-specific genome-scale metabolic modeling (GEM) was generated using the Fast Task-driven Integrative Network Inference for Tissues (ftINIT) algorithm. Further, flux balance analysis (FBA) with ATP hydrolysis as an objective function was performed to study the alterations in metabolic reactions in each sample-specific GEMs.

The Hierarchical and K-means clustering using the matrix representing the presence or absence of metabolic reactions in each GEM identified four clusters of children (After removing an outlier sample). The cluster-specific analysis of FBA results identified reactions related to inflammatory pathways like leukotriene metabolism and omega-3 fatty acid metabolism that are completely absent in cluster 2. In contrast, inflammatory mediators like prostaglandins show very low activation. This indicated the low inflammation characteristic of cluster 2 compared to others.

The low inflammatory characteristic of cluster 2 suggests that the cluster samples have adequate immune systems, making them less susceptible to many diseases. In addition to this, different statistical analyses on the clinical parameters/demographics were done. To check which parameters were significant for which cluster.

Thank you for reading.