PROJECT SUMMARY/ABSTRACT Bacterial vaginosis (BV) is the most common cause for vaginal discharge and highly prevalent among women of reproductive age. The cost of BV sequelae is estimated at $3.7 billion per annum in North America. Microbiologically, BV is characterized by a shift in bacteria from mostly lactobacilli to complex communities of anaerobes. BV is typically diagnosed in the clinic using a set of four clinical findings known as Amsel criteria. The gold standard for BV diagnosis relies on Gram stains of vaginal fluid using Nugent scoring of bacterial morphotypes. Recent advances in molecular detection of bacterial targets have resulted in nucleic acid amplification tests that target combinations of bacteria due to the microbiological heterogeneity of BV. While molecular technologies offer advantages such as detection and quantification of fastidious bacteria, they rely on access to a skilled certified lab and assay turnaround time can delay treatment or result in loss to follow-up. There is a clinical need for a rapid diagnostic that is easy-to-use, sensitive, and specific for BV that can be used in point-of-care (POC) settings. Small molecule metabolites in vaginal fluid have the potential to be developed as biomarkers to be used in rapid diagnostic platforms, but no BV diagnostic products are available using metabolites as biomarkers. We have identified several metabolite biomarkers in preliminary studies conducted in a monomorphic population. The excellent performance of the discovered biomarkers is extremely promising as some individual metabolites have area under the receiving operating curve (AUC) of 95% or higher. We believe it will be possible to combine 2 of these biomarkers to boost this performance even further. A critical next step is to validate these candidate biomarkers in a population that is diverse in ethnicity, racial composition, and economic background to ensure that representative metabolites are selected prior to further product development. In Aim 1, we will validate these biomarkers in 120 women with vaginal symptoms or at high risk for BV. Concentrations of a targeted set of metabolites will be measured by liquid chromatography-mass spectrometry. The diagnostic performance of each individual metabolite will be compared with conventional methods of BV diagnosis, Amsel and Nugent criteria. Metabolites that have the highest diagnostic accuracy will be selected for development of a metabolite-based diagnostic assay for BV in Aim 2. In Aim 2, we will develop and compare the diagnostic performance of a two metabolite BV diagnostic assay with an existing FDA-approved POC that relies on the enzyme sialidase and an FDA-approved molecular test. At the end of STTR Phase I, we expect to have a highly sensitive and specific small molecule metabolite-based BV diagnostic assay that performs as well or outperforms the existing POC, and molecular diagnostic tests based on diagnostic accuracy.