Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis

Abstract Bacterial vaginosis (BV) is a prevalent vaginal syndrome, affecting millions of women globally. The complexity of the vaginal microbiome can challenge conventional diagnostic approaches, particularly for populations of women with healthy, yet diverse vaginal microbiomes. Advanced sequencing...

Full description

Saved in:
Bibliographic Details
Main Authors: Diandra P. Ojo, Cameron Celeste, Dion Ming, Ruogu Fang, Ivana K. Parker
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Women's Health
Online Access:https://doi.org/10.1038/s44294-025-00092-w
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849738374306332672
author Diandra P. Ojo
Cameron Celeste
Dion Ming
Ruogu Fang
Ivana K. Parker
author_facet Diandra P. Ojo
Cameron Celeste
Dion Ming
Ruogu Fang
Ivana K. Parker
author_sort Diandra P. Ojo
collection DOAJ
description Abstract Bacterial vaginosis (BV) is a prevalent vaginal syndrome, affecting millions of women globally. The complexity of the vaginal microbiome can challenge conventional diagnostic approaches, particularly for populations of women with healthy, yet diverse vaginal microbiomes. Advanced sequencing technologies coupled with machine learning (ML) offer promise in elucidating these complexities; however, ML models have been shown to be vulnerable to existing health disparities. To determine the ability of ML models to perform equitably, this study evaluates the performance of ML algorithms in predicting symptomatic BV across different ethnic groups using 16S rRNA sequencing data. Results indicate differential predictive performance across ethnicities, with models exhibiting lower accuracy for Black women. Moreover, we found variation in significant bacterial taxa for predicting BV by ethnicity. Future research aims to explore these factors and validate findings in larger, more diverse cohorts, with the goal of improving BV diagnosis and mitigating health disparities across ethnic groups.
format Article
id doaj-art-d0899f0e05bc43c4800c1110403ff0c2
institution DOAJ
issn 2948-1716
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Women's Health
spelling doaj-art-d0899f0e05bc43c4800c1110403ff0c22025-08-20T03:06:36ZengNature Portfolionpj Women's Health2948-17162025-07-013111010.1038/s44294-025-00092-wPopulation-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosisDiandra P. Ojo0Cameron Celeste1Dion Ming2Ruogu Fang3Ivana K. Parker4J. Crayton Pruitt Family Department of Biomedical Engineering, University of FloridaJ. Crayton Pruitt Family Department of Biomedical Engineering, University of FloridaJ. Crayton Pruitt Family Department of Biomedical Engineering, University of FloridaJ. Crayton Pruitt Family Department of Biomedical Engineering, University of FloridaJ. Crayton Pruitt Family Department of Biomedical Engineering, University of FloridaAbstract Bacterial vaginosis (BV) is a prevalent vaginal syndrome, affecting millions of women globally. The complexity of the vaginal microbiome can challenge conventional diagnostic approaches, particularly for populations of women with healthy, yet diverse vaginal microbiomes. Advanced sequencing technologies coupled with machine learning (ML) offer promise in elucidating these complexities; however, ML models have been shown to be vulnerable to existing health disparities. To determine the ability of ML models to perform equitably, this study evaluates the performance of ML algorithms in predicting symptomatic BV across different ethnic groups using 16S rRNA sequencing data. Results indicate differential predictive performance across ethnicities, with models exhibiting lower accuracy for Black women. Moreover, we found variation in significant bacterial taxa for predicting BV by ethnicity. Future research aims to explore these factors and validate findings in larger, more diverse cohorts, with the goal of improving BV diagnosis and mitigating health disparities across ethnic groups.https://doi.org/10.1038/s44294-025-00092-w
spellingShingle Diandra P. Ojo
Cameron Celeste
Dion Ming
Ruogu Fang
Ivana K. Parker
Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
npj Women's Health
title Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
title_full Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
title_fullStr Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
title_full_unstemmed Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
title_short Population-level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
title_sort population level predictive variation in machine learning diagnosis of symptomatic bacterial vaginosis
url https://doi.org/10.1038/s44294-025-00092-w
work_keys_str_mv AT diandrapojo populationlevelpredictivevariationinmachinelearningdiagnosisofsymptomaticbacterialvaginosis
AT cameronceleste populationlevelpredictivevariationinmachinelearningdiagnosisofsymptomaticbacterialvaginosis
AT dionming populationlevelpredictivevariationinmachinelearningdiagnosisofsymptomaticbacterialvaginosis
AT ruogufang populationlevelpredictivevariationinmachinelearningdiagnosisofsymptomaticbacterialvaginosis
AT ivanakparker populationlevelpredictivevariationinmachinelearningdiagnosisofsymptomaticbacterialvaginosis