Skip to content

This repository contain scripts for ML, single-cell, and proteomics identification of critical regulators of immune cell infiltration and microglia-mediated neuroinflammation following TBI.

Notifications You must be signed in to change notification settings

UFSaidu/Identification_of_TBI_biomarkers_using_Machine_learning

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
 
 
 
 
 
 

Repository files navigation

Traumatic Brain Injury (TBI) Biomarker Discovery

Project Progress

  • Identified 16 key genes from hippocampus bulk microarray data using DESeq2 (DEGs), WGCNA (co-expression networks), PPI (Protein Network Analysis) and GSEA functional enrichment analysis. Genes were primarily inflammatory mediators and involved in inflammation and immune response.
  • Performed feature selection using Lasso, Random Forest (RF), and SVM-RFE on the 16 key genes, identifying 2 core genes as optimal TBI biomarkers.
  • Investigatd diagnostic potential of core genes using GLM, SVM, and XGBoost with ROC analysis.
  • Next, we constructed a nomogram model for risk prediction incorporating core genes.
  • Calibration & Decision Curve Analysis (DCA) were performed to further assess clinical utility of core genes.
  • Immune cell infiltration analysis revealed significant infiltration of neutrophils, monocytes, macrophages, and mast cells following TBI. These cells were positively correlated with core genes, driving neuroinflammation post-injury.
  • single-cell RNA-seq validation of the core genes revealed predominant microglia expression, with minimal expression in astrocytes and endothelial cells. The core genes were involved in leukocyte recruitment and microglia activation, driving neuroinflammation and immune cell infiltration post-TBI.
  • Performed trajectory analysis on microglia subtypes using Slingshot package in R to identify microglia lineages and map pseudotime injury responses.
  • Further validated core genes using microglia RNA-seq data from cortex, incorporating different time-points (acute, subacute, and chronic TBI stages). We observed that core genes exhibited temporal expression patterns across time-points.
  • Age drive TBI progression and outcomes. Thus, immune cell infiltration driven by core genes strongly correlated with age and is associated with injury progression.

Next Steps

  • Proteomic analysis to quantify proteins encoded by the core genes in serum and cerebrospinal fluid.

📌 This README will be updated as the project progresses. Full code will be uploaded to the repository after the paper has been accepted for publication.

Last updated: September 2025

About

This repository contain scripts for ML, single-cell, and proteomics identification of critical regulators of immune cell infiltration and microglia-mediated neuroinflammation following TBI.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages