Skip to content
DraconiticResearch
Search Sources API
190,851 studies31 sourcesUpdated 38 min ago
Peer-reviewedCC BY 4.0PLOS Computational Biology

Predicting bacterial vaginosis incidence using artificial neural networks

  • Jacob H. Elnaggar1
  • John W. Lammons2
  • Caleb M. Ardizzone3
  • Kristal J. Aaron4
  • Clayton Jacobs5
  • Keonte J. Graves6
  • Sheridan D. George7
  • Megan Amerson-Brown
  • Meng Luo
  • Ashutosh Tamhane
  • Paweł Łaniewski
  • Alison J. Quayle
4 more authors
  • Melissa M. Herbst-Kralovetz
  • Nuno Cerca
  • Christina A. Muzny
  • Christopher M. Taylor
7 affiliations
  1. 1Department of Microbiology, Immunology, and Parasitology, Louisiana State University Health Sciences Center, New Orleans, Louisiana, United States of America
  2. 2Division of Infectious Diseases, Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana, United States of America
  3. 3Division of Infectious Diseases, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, United States of America
  4. 4Division of Laboratory Science, Department of Pathology, University of Alabama at Birmingham, Birmingham, Alabama, United States of America
  5. 5Division of Nephrology, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, United States of America
  6. 6Department of Basic Medical Sciences, College of Medicine-Phoenix, University of Arizona, Phoenix, Arizona, United States of America
  7. 7Centre of Biological Engineering, Laboratory of Research in Biofilms Rosário Oliveira, Minho University, Braga, Portugal
Read full textPDFDOICite

Abstract

Background: Bacterial vaginosis (BV) is a vaginal dysbiosis associated with adverse reproductive and infectious outcomes. Current diagnostics identify BV after symptom onset. We evaluated whether artificial neural network (ANN) models of the vaginal microbiome could predict incident BV (iBV) up to 14 days before clinical diagnosis. Methods: ANNs were trained using 16S rRNA gene sequencing data from 1,201 longitudinal vaginal specimens collected from 58 women across two prospective cohorts. Models classified individual specimens as pre-iBV or healthy using the relative abundance of vaginal bacterial taxa. Model performance was assessed using a held-out participant-level test set, participant-level cross-validation, and external validation. SHAP analysis was used to identify taxa associated with model predictions. Results: On the held-out participant-level test set, the ANN achieved 93% accuracy (AUC = 0.97, sensitivity = 95%, specificity = 92%). Models using only five taxa (Lactobacillus crispatus, Gardnerella spp., L. iners, L. mulieris, and Megamonas spp.) maintained >91% accuracy, sensitivity, and specificity. SHAP analysis identified Lactobacillus spp. and Gardnerella spp. as the taxa most strongly associated with model predictions. Participant-level cross-validation produced lower and more variable performance (AUC = 0.826 ± 0.076; accuracy = 71.5% ± 8.2%), while external validation achieved approximately 80% balanced accuracy. Interpretation: Vaginal microbiome composition contains predictive information that can identify women at risk of iBV before clinical onset. Lower performance during participant-level cross-validation and external validation indicates that additional validation in larger, more diverse cohorts is required before clinical implementation. Author summary: Background

Related

  1. Effectiveness of male partner treatment in reducing recurrence of bacterial vaginosis: a systematic review and meta-analysis
    Jeannine Metz, Alex Farr, Herbert Kiss +5
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textMicrobiologyCrossref
  2. Large-scale metagenomic characterization of the vaginal microbiome and uropathogens in women with recent bacterial vaginosis and urinary tract infections
    Sonia N. Whang, Xinyue Wang, Krystal J. Thomas-White +4
    Nature CommunicationsPeer-reviewedCC BY-NC-ND 4.0Full textMicrobiologyCrossref
  3. Exploratory pilot study of a spray-type probiotic delivery system for vaginal health in women at risk for bacterial vaginosis
    Sunwha Park, Juan Kim, Young-Ah You +7
    Scientific ReportsPeer-reviewedCC BY-NC-ND 4.0Full textMicrobiologyCrossref
  4. Effect of pHyph, A novel antibiotic-free treatment, on bacterial vaginosis symptoms and vaginal microbiome composition
    Emilia Lahtinen, Luisa W. Hugerth, Gabriella Edfeldt +4
    European Journal of Obstetrics & Gynecology and Reproductive Biology: XPeer-reviewedCC BY 4.0Full textMicrobiologyPMCEurope PMC
  5. Comparative Performance of Aptima BV/CV-TV and Vaginal Real Time Assays for the Molecular Diagnosis of Bacterial Vaginosis and Vaginitis.
    Laura Gozalo Álvarez, Lucía Pérez Rodríguez, Elvira García-Valdecasas García +3
    Diagnostics (Basel, Switzerland)Peer-reviewedCC BYFull textMicrobiologyEurope PMC

Cite

BibTeXRIS
BibTeX
@article{elnaggar2026predicting,
  title = {{Predicting bacterial vaginosis incidence using artificial neural networks}},
  author = {Jacob H. Elnaggar and John W. Lammons and Caleb M. Ardizzone and Kristal J. Aaron and Clayton Jacobs and Keonte J. Graves and Sheridan D. George and Megan Amerson-Brown and Meng Luo and Ashutosh Tamhane and Paweł Łaniewski and Alison J. Quayle and Melissa M. Herbst-Kralovetz and Nuno Cerca and Christina A. Muzny and Christopher M. Taylor},
  journal = {PLOS Computational Biology},
  year = {2026},
  month = sep,
  doi = {10.1371/journal.pcbi.1014768},
  url = {https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014768},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Predicting bacterial vaginosis incidence using artificial neural networks
AU  - Jacob H. Elnaggar
AU  - John W. Lammons
AU  - Caleb M. Ardizzone
AU  - Kristal J. Aaron
AU  - Clayton Jacobs
AU  - Keonte J. Graves
AU  - Sheridan D. George
AU  - Megan Amerson-Brown
AU  - Meng Luo
AU  - Ashutosh Tamhane
AU  - Paweł Łaniewski
AU  - Alison J. Quayle
AU  - Melissa M. Herbst-Kralovetz
AU  - Nuno Cerca
AU  - Christina A. Muzny
AU  - Christopher M. Taylor
PY  - 2026
DA  - 2026/09/15
JO  - PLOS Computational Biology
DO  - 10.1371/journal.pcbi.1014768
UR  - https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014768
AB  - Background: Bacterial vaginosis (BV) is a vaginal dysbiosis associated with adverse reproductive and infectious outcomes. Current diagnostics identify BV after symptom onset. We evaluated whether artificial neural network (ANN) models of the vaginal microbiome could predict incident BV (iBV) up to 14 days before clinical diagnosis. Methods: ANNs were trained using 16S rRNA gene sequencing data from 1,201 longitudinal vaginal specimens collected from 58 women across two prospective cohorts. Models classified individual specimens as pre-iBV or healthy using the relative abundance of vaginal bacterial taxa. Model performance was assessed using a held-out participant-level test set, participant-level cross-validation, and external validation. SHAP analysis was used to identify taxa associated with model predictions. Results: On the held-out participant-level test set, the ANN achieved 93% accuracy (AUC = 0.97, sensitivity = 95%, specificity = 92%). Models using only five taxa (Lactobacillus crispatus, Gardnerella spp., L. iners, L. mulieris, and Megamonas spp.) maintained >91% accuracy, sensitivity, and specificity. SHAP analysis identified Lactobacillus spp. and Gardnerella spp. as the taxa most strongly associated with model predictions. Participant-level cross-validation produced lower and more variable performance (AUC = 0.826 ± 0.076; accuracy = 71.5% ± 8.2%), while external validation achieved approximately 80% balanced accuracy. Interpretation: Vaginal microbiome composition contains predictive information that can identify women at risk of iBV before clinical onset. Lower performance during participant-level cross-validation and external validation indicates that additional validation in larger, more diverse cohorts is required before clinical implementation. Author summary: Background
SN  - 1553-7358
ER  -