Abstract
Research on the human gut microbiome is expanding rapidly; yet, most published studies focus on populations from high-income regions such as North America and Europe. Underrepresentation of populations from low- and middle-income countries in the microbiome literature limits the generalizability of microbiome-health associations. These challenges are compounded by computational barriers, including biases in reference databases, nonrepresentative metadata, and infrastructure limitations in low- and middle-income countries. However, recent efforts in large-scale global sampling have begun to address these problems. This review provides recommendations for future research efforts applying computational analysis to global microbiome data, including guidelines to initiate and maintain equitable partnerships, identify representative datasets, overcome technical limitations, and contextualize results at the global scale.