Sarra4299/pubmed-microbiome-mena

Literature mining analysis: quantifying the microbiome research gap in MENA/North African populations. Uses PubMed API + NLP to identify publication trends, research gaps, and disease focus areas. # PubMed Microbiome MENA Literature-mining project quantifying the microbiome research gap between MENA / North African and European populations, using the PubMed E-utilities API. > **Status: work in progress.** The data collection pipeline is complete and produces the dataset described below. The analysis stage is in development. See Roadmap. ## Motivation Human microbiome composition varies substantially with diet, geography, and ancestry meaning findings from one population do not transfer automatically to another. Reference datasets and disease associations built mainly on European and North American cohorts may therefore describe populations that are not representative of most of the world. This project measures that imbalance directly for the MENA / North African region rather than assuming it: how much microbiome literature exists per country, how it has grown over time, and which diseases it covers. ## Research Question This project investigates whether microbiome research on MENA/North African populations is underrepresented compared to European populations, and whether this gap is closing over time. Specifically, we examine the volume of published microbiome literature for three North African countries (Tunisia, Morocco, Egypt) and compare it to two European Mediterranean countries (Italy, France). We further examine which disease areas are covered in each region, in order to identify not only how much research exists, but which topics remain understudied. Three questions are addressed: How large is the difference in publication volume between the MENA and European countries? How has publication volume in each region developed over time? Which disease areas are represented in the MENA literature, and which are missing relative to Europe? ## Data collection Publication records are retrieved from PubMed via the NCBI E-utilities API in a two-step process: 1. **`esearch`** :retrieve PMIDs matching a search term (up to 200 per country) 2. **`efetch` …