chrislubega/Gut_microbial_profiling_of_COVID-19_Patients_in_Uganda

This repository contatins analysis scripts for the paper titled Gut Microbial Profiling of COVID-19 patients in Uganda # Gut Microbiome Analysis of Ugandan COVID-19 Patients ## Overview Analysis pipeline for a study examining gut microbiome composition and its association with COVID-19 severity in 100 confirmed cases from Kampala, Uganda (2020–2021 cohorts). ## My Role Bioinformatics analyst: received raw sequencing data and demographic data from the PI, and independently designed and executed the full analytical pipeline, from raw reads to published results. ## Methods - **Sequencing**: 16S rRNA gene sequencing on DNA extracted from stool samples - **Taxonomic Classification**: Kraken2 — k-mer based comparison of sequence fragments against a reference database to classify microbial composition per sample - **Statistics**: Comparative analysis of microbial composition across COVID-19 status and severity groups - **Machine Learning**: Compared four models — Logistic Regression, Gradient Boosting, Neural Networks, and Random Forest — to identify microbial and demographic predictors of disease severity ## Key Findings - Identified elevated pathogenic taxa (*Klebsiella oxytoca*, *Salmonella enterica*, *Serratia marcescens*) in COVID-19 patients, particularly severe cases - Found reduced beneficial microbes (e.g. Alphaproteobacteria) associated with severity - ML models identified age and specific taxa (Ruminococcaceae, Bacilli, Enterobacteriales, Porphyromonadaceae, *Prevotella copri*) as predictive of severity ## Citation Kateete DP, Lubega C, Nasinghe E, Mbabazi M, Galiwango R, Jjingo D. Gut microbial profiles of COVID-19 patients in Uganda. Afr Health Sci. 2026 Mar;26(1):1-15. doi: 10.4314/ahs.v26i1.2. PMID: 42063908.