20The vaginal microbiome plays an influential role in several disease states in reproductive age 21 women, including bacterial vaginosis (BV). While demographic characteristics are associated 22 with differences in vaginal microbiome community structure, little is known about the influence 23 of sexual and hygiene habits. Furthermore, associations between the vaginal microbiome and risk 24 symptoms of bacterial vaginosis have not been fully elucidated. Using Bayesian network (BN) 25 analysis of 16S rRNA gene sequence results, demographic and extensive questionnaire data, we 26 describe both novel and previously documented associations between habits of women and their 27 vaginal microbiome. The BN analysis approach shows promise in uncovering complex 28 associations between disparate data types. Our findings based on this approach support published 29 associations between specific microbiome members (e.g., Eggerthella, Gardnerella, Dialister, 30 . CC-BY-NC-ND 4.0 International license It is made available under a was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. indicating that the Nugent Score may not be the most useful criteria for assessment of clinical 36 BV. We also found that demographics (i.e., age, ethnicity, previous pregnancy) were associated 37 with the presence/absence of specific vaginal microbes. The resulting BN revealed several as-yet 38 undocumented associations between birth control usage, menstrual hygiene practices and 39 specific microbiome members. Many of these complex relationships were not identified using 40 common analytical methods, i.e., ordination and PERMANOVA. While these associations 41 require confirmatory follow-up study, our findings strongly suggest that future studies of the 42 vaginal microbiome and vaginal pathologies should include detailed surveys of participants' 43 sanitary, sexual and birth control habits, as these can act as confounders in the relationship 44 between the microbiome and disease. Although the BN approach is powerful in revealing 45 complex associations within multidimensional datasets, the need in some cases to discretize the 46 data for use in BN analysis can result in loss of information. Future research is required to 47 alleviate such limitations in constructing BN networks. Large sample sizes are also required in 48 order to allow for the incorporation of a large number of variables (nodes) into the BN, 49 particularly when studying associations between metadata and the microbiome. We believe that 50 this approach is of great value, complementing other methods, to further our understanding of 51 complex associations characteristic of microbiome research. 52 53. CC-BY-NC-ND 4.0 International license It is made available under a was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.

The mcaGUI package and source are freely available as part of Bionconductor at http://www.bioconductor.org/packages/release/bioc/html/mcaGUI.html