Résumé
Background The composition of the vaginal microbiota is known to be highly structured into five main community state types (CSTs) that are found in all human populations. Several associations between perceived ethnicity and the type of community have been reported but analyses of human genetic data, especially genome wide association studies (GWAS), remain limited and mostly rely on phenotypic traits rather than microbial DNA data. Methods Analysing genotyping data from in 168 women from the PAPCLEAR cohort study in France, we perform a GWAS looking for human genetic polymorphisms associated with vaginal microbiota community composition. For the latter, we use Simpson diversity and community state type (CST) as summary statistics to summarise 16S RNA metabarcoding data. Results We show that inverse Simpson diversity is the trait related to the vaginal microbiota that is best explained by the human genome. Furthermore, we identify several genomic regions associated with variations in this trait and show that the covariates associated with vaginal microbiota composition do not correlate with these genetic variants. Conclusion This is one of the first GWAS to use microbial genetic data instead of symptoms to characterise the vaginal microbiota. However, it remains limited because of the size of our cohort and our results call for more powered studies in terms of participants and genome coverage.