Microbiome-based assessment and treatment of diseases and disorders

Abstract

Methods, systems, computer-readable media, and kits are provided for assessing and managing the risk of diseases or disorders using molecular cervicovaginal microbiome features derived from bacterial 16S rRNA gene sequencing. A cervicovaginal microbiome profile, including a molecular bacterial vaginosis (molBV) score, is determined. A subject is classified as being at elevated risk if the profile indicates an mBV-positive state with an elevated level of Candidatus Lachnocurva vaginae or if a microbial risk score (MRS) exceeds a predetermined threshold.

Description

[0001] Docket No. 182219.00297

[0002] MICROBIOME-BASED ASSESSMENT AND TREATMENT OF DISEASES AND DISORDERS

[0003] CROSS-REFERENCE TO RELATED APPLICATIONS This application is entitled to priority pursuant to 35 U. S. C. § 119(e) to U. S. Provisional Patent Application No, 63 / 745,348, filed on January 15, 2025. The content of the application is incorporated herein by reference in its entirety.

[0004] REFERENCE TO AN ELECTRONIC SEQUENCE LISTING

[0005] The contents of the electronic sequence listing (182219.00297SeqList. xml; Size: 4,566 bytes; and Date of Creation: December 28, 2025) are herein incorporated by reference in its entirety.

[0006] FIELD OF THE INVENTION

[0007] This disclosure relates generally to kits and methods for microbiome-based assessment and treatment of diseases or disorders.

[0008] BACKGROUND OF THE INVENTION

[0009] The cervicovaginal microbiome (CVM) comprises complex bacterial and fungal communities that play a central role in maintaining genital tract homeostasis, mucosal immunity, and reproductive health. CVM community structures are commonly categorized by dominance of Lactobacillus species (including L. crispatus, L. iners, L. gasseri, and L. jensenii) or by polymicrobial states characterized by reduced Lactobacillus abundance and increased microbial diversity. Lactobacillus-dominant communities are generally associated with low vaginal pH and protection against pathogens through lactic acid production and additional community -level mechanisms. In contrast, polymicrobial communities are associated with increased inflammation and heightened susceptibility to adverse gynecologic and reproductive outcomes.

[0010] Bacterial vaginosis (BV) represents the most common form of vaginal dysbiosis worldwide. BV is traditionally defined by depletion of protective Lactobacillus species and overgrowth of anaerobic taxa. Clinical diagnosis relies primarily on Amsel criteria or Nugent scoring, both of which collapse biologically diverse microbial communities into a single diagnostic category. These approaches fail to capture the substantial molecular heterogeneity of BV and provide limited insight into prognosis, recurrence risk, or disease susceptibility. AsDocket No. 182219.00297

[0011] a result, patients classified as “BV positive” may harbor fundamentally different microbial states with distinct biological consequences.

[0012] Emerging molecular studies demonstrate that BV is not a single entity but rather comprises distinct molecular subtypes characterized by reproducible microbial signatures. These subtypes differ in dominant taxa, stability over time, inflammatory potential, and clinical risk profiles. However, there remains no standardized framework for molecular BV subtyping that can be implemented clinically to stratify patients by risk or guide management decisions. This represents a critical unmet need in women’s health.

[0013] Disruptions to the CVM, and specifically to defined BV molecular subtypes, have been associated with increased risk for a wide range of conditions, including sexually transmitted infections, pelvic inflammatory' disease, adverse pregnancy outcomes, and cervical disease. For example, BV has been consistently associated with preterm birth, yet only a subset of BV-positive individuals experience adverse pregnancy outcomes, suggesting that specific microbial configurations drive risk. Similarly, persistent dysbiosis has been implicated in cervical intraepithelial neoplasia (CIN) progression, where inflammatory and anaerobe-rich communities may promote viral persistence and epithelial transformation.

[0014] Alterations in the CVM have also been observed following pelvic radiation therapy, where treatment-induced mucosal injury and immune disruption lead to long-lasting microbiome shifts. These post-radiation microbiome changes are increasingly recognized as contributors to chronic symptoms, infection susceptibility, and impaired tissue recovery, yet no tools currently exist to stratify patients based on molecular CVM patterns following treatment.

[0015] Among infectious diseases influenced by CVM structure, Chlamydia trachomatis (C T) represents a highly prevalent and clinically consequential example. CT is frequently asymptomatic but can lead to pelvic inflammatory disease, chronic pelvic pain, ectopic pregnancy, and infertility. Perinatal transmission may result in neonatal conjunctivitis and pneumonia. Incident infections disproportionately affect adolescents and young adults, with pronounced racial, ethnic, and socioeconomic disparities.

[0016] Epidemiologic studies have identified behavioral risk factors for CT acquisition, including younger age, multiple sexual partners, and prior sexually transmitted infections. BV has also been proposed as a biological risk factor. However, traditional BV diagnostics fail to distinguish which individuals are truly at elevated risk, as they do not account for molecularDocket No. 182219.00297

[0017] subtype heterogeneity. This obscures causal inference and limits the ability to implement targeted prevention strategies.

[0018] Multiple investigations have examined associations between CVM composition and subsequent CT detection. Nevertheless, the specific molecular features and temporal dynamics of the microbiome that precede incident infection remain incompletely characterized. Importantly, existing studies largely treat BV as a binary state rather than a spectrum of biologically distinct subtypes with differential risk profiles.

[0019] Accordingly, there is a need for microbiome-based methods that stratify BV into molecular subtypes using defined microbial markers. Such approaches would enable prospective risk assessment, integration of relevant clinical and behavioral covariates, and improved clinical decision-making. Molecular subtyping of BV has the potential to identify individuals at highest risk for CT infection, CIN progression, preterm birth, post-radiation complications, and other adverse outcomes, thereby enabling personalized surveillance and microbiome-targeted interventions to reduce morbidity.

[0020] SUMMARY OF THE INVENTION

[0021] This disclosure provides methods, systems, and kits for assessing a subject’s risk of a disease or disorder by analyzing their cervicovaginal microbiome (CVM). The methods generally involve obtaining a vaginal or cervicovaginal sample from the subject and determining a CVM profile, which includes a molecular bacterial vaginosis (mBV) score (molBV) and quantification of cornerstone vaginal bacterial species (i.e., Candidatus Lachnocurva vaginae, also known as BVAB1). The subject is then classified as having an elevated risk based on specific microbiome signatures, including identification of an mBV-A subtype defined by (i) an mBV-positive state (e.g., molBV > 7) and elevated abundance of Candidatus Lachnocurva vaginae relative to a predetermined threshold, or (ii) a microbial risk score (MRS) that exceeds a predetermined threshold.

[0022] Also provided are methods of treating subjects identified as having an elevated risk, which may include administering microbiome-targeted therapies. The disclosure further includes computer-implemented systems configured to perform the risk assessment by processing microbiome data and kits containing reagents for sample analysis. These methods, systems, and kits are applicable to assessing risk for and guiding management of a range of diseases and disorders, including but not limited to Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitalium infection;Docket No. 182219.00297

[0023] pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

[0024] In some implementations, the method comprises non-routine laboratory processing of clinical cytology specimens, computational transformations specific to compositional microbiome data (including centered log-ratio transformation, optional trained nearest-centroid community state-type (CST) classification, and calibrated microbial risk scoring), and automatically generating a machine-readable report configured to trigger a predefined clinical workflow, such as scheduling follow-up testing or initiating microbiome-targeted therapy per protocol.

[0025] In one aspect, this disclosure provides an in vitro method of assessing risk of a disease or disorder in a subject with a vagina or cervix or formerly with a cervix. In some embodiments, the method comprising: (a) obtaining a vaginal or cervicovaginal sample from the subject; (b) determining, from the sample, a cervicovaginal microbiome profile comprising at least: a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology, including shotgun metagenomics or targeted quantitative PCR (qPCR) of one or more key bacterial species, including Candidatus Lachnocurva vaginae (c) classifying the subject as having an elevated risk of disease or disorder when the microbiome profile indicates (i) an mBV-positive state and an elevated level of Candidatus Lachnocurva vaginae, or (ii) a microbial risk score (MRS) above a predetermined threshold derived from abundances of a plurality of B V -associated taxa.

[0026] In another aspect, this disclosure also provides a method of treating a disease or disorder in a subject with a vagina or cervix. In some embodiments, the method comprising: (a) obtaining a vaginal or cervicovaginal sample from the subject; (b) determining, from the sample, a cervicovaginal microbiome profile comprising at least: a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology, including shotgun metagenomics or targetedDocket No. 182219.00297

[0027] quantitative PCR (qPCR) of one or more key bacterial species, including Candidatus Lachnocurva vaginae; (c) classifying the subject as having an elevated risk of disease or disorder when the microbiome profile indicates (i) an mBV-positive state and an elevated level of Candidatus Lachnocurva vaginae, or (ii) a microbial risk score (MRS) above a predetermined threshold derived from abundances of a plurality of BV-associated taxa; and (d) when the classification indicates elevated risk, administering a therapy according to a dosing regimen effective to reduce a risk of the disease or disorder in the subject.

[0028] In some embodiments, the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders (e.g., Lupus Erythematosus).

[0029] In some embodiments, the therapy comprising a microbiome-targeted therapy that comprises vaginal Lactobacillus probiotic therapy; a non-antibiotic biofilm-disrupting agent; and / or a prebiotic therapy facilitating the growth of Lactobacillus or other healthy bacteria.

[0030] In some embodiments, the method comprises computing the microbial risk score by: performing amplicon sequence variant inference on 16S reads; assigning taxonomy with a trained classifier; transforming taxa abundances via centered log-ratio; computing log-ratios relative to Lactobacillus; and applying pre-trained weights with Platt-scaled calibration to output a probability of incident infection.

[0031] In some embodiments, taxa abundances used for microbial risk scoring are obtained by targeted quantitative PCR (qPCR), digital PCR, hybrid capture, or other nucleic acid-based quantification methods.

[0032] In some embodiments, the microbiome profile further comprises a ratio of Candidatus Lachnocurva vaginae to Lactobacillus, and the subject is classified as elevated risk when the ratio exceeds a predetermined cutoff.Docket No. 182219.00297

[0033] In some embodiments, the plurality of B V-associated taxa used to compute the microbial risk score comprises at least Candidatus Lachnocurva vaginae and one or more of Prevotella, Megasphaera, Clostridium, Staphylococcus, Acinetobacter, or additional taxa selected by differential abundance analysis and cross-validation.

[0034] In some embodiments, the method further comprises outputting a report comprising the subject’s risk classification, the mBV category, and the microbial risk score value, and optionally a CST assignment.

[0035] In some embodiments, the risk assessment is prospective, based on a sample collected prior to detection of incident infection.

[0036] In some embodiments, the method assesses risk of reinfection following antibiotic treatment, based on a post-treatment sample.

[0037] In some embodiments, the CST assignment is performed by a nearest-centroid classification model trained on cervi co vaginal community compositions.

[0038] In some embodiments, the microbial risk score is a weighted sum of taxa abundances or log-ratios relative to Lactobacillus and the weights are derived from odds ratios for incident infection. In some embodiments, the microbial risk score is scaled and categorized using predetermined cutoffs corresponding to increasing odds of incident infection.

[0039] In some embodiments, the method comprises selecting a panel of BV-associated taxa by differential abundance analysis with false-disco very -rate control and cross-validation; and transforming abundances by CLR; compute log-ratios of selected taxa relative to Lactobacillus group abundance.

[0040] In some embodiments, the method further comprises integrating a sexual risk behavior score and / or high-risk HPV status as covariates into the risk classification model.

[0041] In some embodiments, the classification step is performed by a trained machinelearning model that outputs a categorical risk label and / or a probability score. In some embodiments, the classifier is calibrated such that mBV-intermediate yields an intermediate risk categoiy and mBV-A yields a higher risk category.

[0042] In some embodiments, the method further comprises recommending an intervention selected from follow-up testing, antibiotic stewardship, microbiome-targeted therapy, prebiotic or probiotic therapy, or behavioral counseling based on the elevated risk classification.Docket No. 182219.00297

[0043] In some embodiments, the sample is a cervical cytology specimen collected in a liquid-based cytology medium. In some embodiments, the analysis is performed on archived DNA from a Pap-smear specimen.

[0044] In some embodiments, the method further comprises detecting high-risk HPV genotypes and reporting the combined microbiome and HPV risk profile.

[0045] In some embodiments, the mBV-A subtype is defined by an mBV-positive state combined with elevated abundance of Candidatus Lachnocurva vaginae, and the subject is classified as elevated risk upon identification of mBV-A.

[0046] In some embodiments, the elevated risk classification triggers a programmed testing schedule with an increased surveillance interval.

[0047] In some embodiments, the method outputs a longitudinal comparison between a preinfection sample and a post-treatment sample, highlighting persistence or emergence of the mBV-A subtype.

[0048] In some embodiments, the pipeline flags subjects exhibiting a transition toward mBV- positive states as at risk of incident infection.

[0049] In another aspect, this disclosure provides a method of guiding clinical management for a disease or disorder. In some embodiments, the method comprises: performing the method as described herein to identify a subject at elevated risk of the disease or disorder; and recommending at least one of microbiome-targeted therapy, probiotic therapy, or adjusted follow-up intervals.

[0050] In yet another aspect, this disclosure further provides a computer-implemented system for assessing risk of a disease or disorder in a subject with a vagina and / or cervix. In some embodiments, the system comprises: (a) a data processor and memory storing instructions that, when executed, cause the system to: receive subject-level microbiome data comprising bacterial 16S rRNA gene sequencing features from a vaginal or cervicovaginal sample; compute a molBV score and optionally assign an mBV subtype; and compute a microbial risk score from a set of BV-associated taxa; and classify a risk of disease or disorder, such as Chlamydia trachomatis infection, in the subject based on: (i) mBV-positive with an elevated level of Candidatus Lachnocurva vaginae, or (ii) the microbial risk score exceeding a threshold; and (b) an interface configured to generate a report comprising the risk classification.Docket No. 182219.00297

[0051] In some embodiments, the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM ); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

[0052] In some embodiments, the instructions further cause the system to compute a ratio of Candidatus Lachnocurva vaginae to Lactobacillus and incorporate the ratio into the risk classification.

[0053] In some embodiments, the microbial risk score is calculated as a weighted sum of abundances or log-ratios of taxa relative to Lactobacillus, with weights derived from effect sizes for incident infection.

[0054] In some embodiments, the system further comprises a model-training module configured to update classification thresholds or weights based on newly ingested labeled data.

[0055] In another aspect, this disclosure provides a non-transitory computer-readable medium storing instructions that, when executed by a processor, cause a computing device to perform the method as described herein to identify a subject at elevated risk of the disease or disorder.

[0056] In another aspect, this disclosure additionally provides a kit for assessing risk of a disease or disorder in a subject with a vagina and / or cervix. In some embodiments, the kit comprises: (a) reagents for extraction of nucleic acids from a vaginal or cervicovaginal sample; (b) a primer set configured to amplify bacterial 16S rRNA gene sequences from the sample; (c) optionally, a primer set configured to amplify fungal or eukaryotic ITS sequences from the sample; (d) one or more positive controls and negative controls for sequencing and analysis quality assurance; and (e) instructions to access software or an analysis pipeline configured to compute an molBV, optionally assign a CST, compute a microbial risk score from BV-associated taxa, and classify infection risk based on: (i) mBV-positive with an elevated level of Candidatus Lachnocurva vaginae, or (ii) the microbial risk score.Docket No. 182219.00297

[0057] In some embodiments, the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM ); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

[0058] In some embodiments, the kit further comprises a sample collection device and transport medium suitable for preserving cervicovaginal specimens for microbiome analysis.

[0059] In some embodiments, the software implements a nearest-centroid classifier to optionally assign CSTs and a trained risk model that outputs a risk label or probability for incident or recurrent infection. In some embodiments, the software reports a ratio of Candidatus Lachnocurva vaginae to Lactobacillus and compares the ratio to a predetermined cutoff associated with elevated risk. In some embodiments, the software enables integration of covariates comprising sexual risk behavior metrics and high-risk HPV status into the risk classification.

[0060] In some embodiments, the kit further comprises written instructions indicating that elevated risk is determined upon identification of an mBV-positive state with an elevated level of Candidatus Lachnocurva vaginae or a microbial risk score above a threshold, and recommending confirmatory testing or clinical follow-up.

[0061] In some embodiments, the positive control comprises a defined mock community including Lactobacillus and selected BV-associated taxa, and the negative control comprises nuclease-free water. In some embodiments, the positive control comprises a defined mock community including Lactobacillus crispatus, Lactobacillus iners, Candidatus Lachnocurva vaginae, and Prevotella bivia.

[0062] In some embodiments, the analysis pipeline computes the molBV score by transforming 16S rRNA gene sequencing data into a Nugent-like score on a 0-10 scale and classifies the score into negative, intermediate, or positive categories. In some embodiments, the pipelineDocket No. 182219.00297

[0063] produces both prospective risk for incident infection and risk of reinfection following treatment, based on pre- and post-treatment sampling.

[0064] In some embodiments, the microbial risk score is derived from a validated set of BV-associated taxa including at least Candidatus Lachnocurva vaginae and Prevotella.

[0065] In some embodiments, the primer set targets the V1-V3, V3-V4, V4, V4-V5, or V5-V7 region of the bacterial 16S rRNA gene.

[0066] BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 illustrates an overview of the prospective CT natural history study design. Case subjects (n = 187) were selected based on identification of an incident CT infection diagnosed during routine clinical screening using the Gen-Probe APTIMA test. Controls (n = 373) were matched to cases at a 2:1 ratio at the time of incident CT infection, based on age, date of study enrollment, and prior history of CT infection. Retrospective cervicovaginal swab samples were retrieved from prior visits for all participants (approximately six months for both cases and controls). Post-treatment samples were collected from 504 of 560 participants (approximately eight months after treatment of the case CT infection). All cervical samples underwent 16SV4 rRNA and ITS1 amplicon sequencing to assess bacteria and fungi / eukaryotes, respectively. HPV genotyping was performed using data from the parent study and included high-risk (oncogenic) HPV types: HPV16, HPV18, HPV31, HPV33, HPV35, HPV39, HPV45, HPV51, HPV52, HPV56, HPV58, and HPV59. Multivariate modeling was used to determine (1) the prospective risk of CT acquisition (at t-1), (2) the association of CT infection with the cervicovaginal microbiome (CVM) (at tO), and (3) residual effects of treated CT infection on the CVM (at t+1).

[0067] FIG. 2 shows the cross-validated combined common effects of all bacterial genera identified as significantly associated with prospective CT acquisition by ANCOM. Of the 18 genera identified, 10 were significant in the validation phase.

[0068] FIG. 3 is a Venn diagram showing mBV-A subtype overlap between t-i and t+j visits in participants with post-t+1 CT follow-up data, related to Table 3, model B. The Venn diagram shows the number of participants with the mBV-A subtype at the t-1 and t+1 visits, as well as the number of participants that had this mBV subtype in both visits.

[0069] DETAILED DESCRIPTION OF THE INVENTIONDocket No. 182219.00297

[0070] Methods, systems, computer-readable media, and kits are provided for assessing and managing risk of diseases or disorders in subjects born with a vagina and / or cervix or formerly possessing a cervix using cervicovaginal microbiome features derived from bacterial 16S rRNA gene sequencing or other nucleic acid-based technologies. A cervicovaginal microbiome profile comprising a molecular bacterial vaginosis (molBV) score equivalent to mBV-positive and elevated abundance of Candidatus Lachnocurva vaginae is determined, and the subject is classified as elevated risk when the profile indicates an mBV-positive state with elevated Candidatus Lachnocurva vaginae (defined as mBV-A) or a microbial risk score (MRS) exceeds a threshold. The MRS can be computed by centered log-ratio transformation, log-ratios relative to Lactobacillus, and application of pre-trained weights with Platt-scaled calibration; taxa can include Candidatus Lachnocurva vaginae and Prevotella. Methods can prospectively predict incident infection, assess reinfection post-treatment, and integrate covariates such as sexual risk behavior and high-risk HPV status. In some embodiments, elevated risk triggers administration of a microbiome-targeted therapy and / or adjusted follow-up. Kits provide reagents, controls, and software implementing the analysis.

[0071] In some implementations, the method comprises non-routine laboratory processing of clinical cytology specimens, computational transformations specific to compositional microbiome data (including centered log-ratio transformation, trained nearest-centroid CST classification, and calibrated microbial risk scoring), and automatically generating a machine- readable report configured to trigger a predefined clinical workflow, such as scheduling follow¬ up testing or initiating microbiome-targeted therapy per protocol.

[0072] In one aspect, this disclosure provides methods, systems, and kits for assessing a subject’s risk of a disease or disorder by analyzing their cervicovaginal microbiome (CVM). The methods generally involve obtaining a vaginal or cervicovaginal sample from the subject and determining a CVM profile, which includes a molecular bacterial vaginosis (molBV) score and a community state type (CST) assignment. The subject is then classified as having an elevated risk based on specific microbiome signatures, such as the presence of an mBV-A subtype (defined by an mBV-positive state with an elevated level of Candidatus Lachnocurva vaginae) or a microbial risk score (MRS) that exceeds a predetermined threshold.

[0073] Also provided are methods of treating subjects identified as having an elevated risk, which may include administering microbiome-targeted therapies. The disclosure further includes computer-implemented systems configured to perform the risk assessment byDocket No. 182219.00297

[0074] processing microbiome data and kits containing reagents for sample analysis. These methods, systems, and kits are applicable to assessing risk for and guiding management of a range of diseases and disorders, including Chlamydia trachomatis infection, other sexually transmitted infections (STIs), pelvic inflammatory disease (PID), and adverse pregnancy outcomes.

[0075] Methods for Diagnosing and Treating Diseases or Disorders

[0076] In one aspect, this disclosure provides an in vitro method of assessing risk of a disease or disorder in a subject with a vagina and / or cervix. In some embodiments, the method comprising: (a) obtaining a vaginal or cervi covagin al sample from the subject; (b) determining, from the sample, a cervicovaginal microbiome profile comprising at least a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology; and (c) classifying the subject as having an elevated risk of disease or disorder when the microbiome profile indicates (i) an mBV-positive state and an elevated level of Candidatus Lachnocurva vaginae, or (ii) a microbial risk score (MRS) above a predetermined threshold derived from abundances of a plurality of BV-associated taxa.

[0077] In another aspect, this disclosure also provides a method of treating a disease or disorder in a subject with a vagina and / or cervix. In some embodiments, the method comprising: (a) obtaining a vaginal or cervicovaginal sample from the subject; (b) determining, from the sample, a cervicovaginal microbiome profile comprising at least a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology; Candidatus Lachnocurva vaginae levels; (c) classifying the subject as having an elevated risk of disease or disorder when the microbiome profile indicates (i) an mBV-positive state and Candidatus Lachnocurva vaginae (and optionally a CST-IV-A status), or (ii) a microbial risk score (MRS) above a predetermined threshold derived from abundances of a plurality of BV-associated taxa; and (d) when the classification indicates elevated risk, administering a therapy according to a dosing regimen effective to reduce a risk of the disease or disorder in the subject.

[0078] As used herein, the term “subject with a cervix” refers to a human subject possessing a cervix at birth, including but not limited to cisgender women, transgender men with a cervix, post-hysterectomy cisgender women and nonbinary individuals formerly possessing a cervix, unless otherwise indicated by context.Docket No. 182219.00297

[0079] As used herein, “Pelvic Inflammatory Disease (PID)” refers to an infection of the female reproductive organs, including the uterus, fallopian tubes, and ovaries. It is often a complication of a sexually transmitted infection.

[0080] As used herein, a “subject” or “individual” means a human or animal. Usually, the animal is a vertebrate such as a primate, rodent, domestic animal or game animal. Primates include chimpanzees, cynomolgus monkeys, spider monkeys, and macaques, e.g., Rhesus. Rodents include mice, rats, woodchucks, ferrets, rabbits and hamsters. Domestic and game animals include cows, horses, pigs, sheep, goats, deer, bison, buffalo, feline species, e.g., domestic cat, canine species, e.g., dog, fox, wolf, avian species, e.g., chicken, emu, ostrich, and fish, e.g., trout, catfish and salmon. In some embodiments, the subject is a mammal, e.g., a human or a non-human mammal. The mammal can be a human, non-human primate, mouse, rat, dog, cat, horse, or cow, but is not limited to these examples. Mammals other than humans can be advantageously used as subjects that represent animal models of disorders. The terms, “individual,” “patient” and “subject” are used interchangeably herein. In some embodiments, the subject is female.

[0081] As used herein, the term “treating” or “treatment” of any disease or disorder refers in one embodiment, to ameliorating a disease or disorder (i.e., arresting or reducing the development of the disease or at least one of the clinical symptoms thereof). In another embodiment, “treating” or “treatment” refers to ameliorating at least one physical parameter, which may not be discernible by the patient. In yet another embodiment, “treating” or “treatment” refers to modulating the disease or disorder, either physically, (e.g., stabilization of a discernible symptom), physiologically, (e.g., stabilization of a physical parameter), or both. In yet another embodiment, “treating” or “treatment” refers to preventing or delaying the onset or development or progression of the disease or disorder.

[0082] As used herein, the term “cervicovaginal sample” refers to a vaginal or cervicovaginal sample and encompasses any specimen containing secretions, cells, and / or microbiota from the vagina or cervix. Examples include endocervical swabs, vaginal swabs, cervical cytology specimens collected in liquid-based cytology media (e.g., ThinPrep®, SurePath®), clinician-or self-collected swabs, tampons, sponges, or lavage fluids. Samples may be fresh, frozen, or preserved in nucleic acid stabilization media. ThinPrep is a registered trademark of Hologic, Inc.; SurePath is a registered trademark of Becton, Dickinson and Company. The use of third-party marks herein is for identification only and does not imply affiliation or endorsement.Docket No. 182219.00297

[0083] In some embodiments, a clinician- or self-collected endocervical or vaginal swab is obtained and placed in a nucleic acid preservation solution. Alternatively, a cervical cytology specimen collected in a liquid-based cytology medium is used, including archived DNA from a Pap-smear specimen. The sample is vortexed and aliquoted for DNA extraction using silica column or magnetic bead-based methods. DNA extraction includes negative controls (e.g., nuclease-free water) and positive controls (defined mock communities) to monitor contamination and batch effects. Extracted nucleic acids are quantified and used to prepare 16S rRNA gene amplicon libraries targeting, for example, the V1-V3, V3-V4, V4, V4-V5, or V5-V7 region using primers with sample-specific barcodes and Illumina-compatible adapters. Library QC includes fragment size verification and quantification; pooled libraries are sequenced to a sufficient depth to resolve the relative abundances of target taxa (e.g., median 10,000-50,000 reads / sample), acknowledging that precise depth may vary by platform and performance needs. Optional ITS amplification is performed for fungal profiling, using standardized ITS1 or ITS2 primer sets.

[0084] In some embodiments, samples failing quality thresholds, such as bacterial 16S read depth <10,000 reads, negative control recovery exceeding 5% of median true sample reads, or positive control composition deviating by >2 standard deviations from expected — are flagged, withheld from classification, and recommended for re-extraction or resequencing per kit instructions.

[0085] As used herein, the term “microbiome profile” refers to a set of quantitative and / or categorical features characterizing the cervi covagin al microbiota. The profile includes at least a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology and Candidatus Lachnocurva vaginae quantification using 16S, qPCR or similar nucleic acid approach. Optionally, the profile may include one or more of a microbial risk score (MRS), ratios of selected taxa (e.g., Prevotella to Lactobacillus), absolute or relative abundances, diversity indices, CST (particularly CST-IV-A measurement), and derived log-ratios.

[0086] As used herein, an “mBV-positive state” refers to a condition in which the molecular bacterial vaginosis (molBV) score is greater than, for example, 7 (on a 0-10 scale), as determined by a validated molecular diagnostic assay. The molBV score is typically calculated based on the relative abundance of specific bacterial taxa associated with bacterial vaginosis, such as Gardnerella vaginalis, Atopobium vaginae, and other anaerobic species, compared toDocket No. 182219.00297

[0087] lactobacilli species. The assay may employ quantitative PCR, next-generation sequencing, or other nucleic acid-based techniques to measure bacterial composition and assign a numerical score according to a predefined algorithm. A molBV score greater than 7 indicates a significant shift in the vaginal microbiome consistent with bacterial vaginosis using the system of Nugent scores as a clinical reference. Thus, mBV is the categorical or ordinal state derived from the molBV score.

[0088] As used herein, the term “molecular bacterial vaginosis (molBV) score” denotes a score derived from 16S rRNA gene sequencing features that approximates a Nugent -like scale (e.g., 0-10) and is categorized as negative, intermediate, or positive according to predetermined thresholds. In some embodiments, the molBV score is generated by transforming amplicon sequence variant (ASV) features associated with Lactobacillus and BV-associated taxa into a weighted or rule-based index.

[0089] In some embodiments, ASV features associated with Lactobacillus and BV-associated taxa are mapped to a Nugent-like 0-10 index as follows: (a) construct scores S Lacto and S BV by summing or weighting CLR-transformed abundances of Lactobacillus and BV-associated taxa, respectively; (b) Transform S Lacto and S BV to percentile ranks or z-scores, and compute molBV = f(S_BV ~ S_Lacto), where f maps the continuous difference to a 0-10 scale using linear or monotonic spline calibration learned from reference data; and (c) categorize as mBV-negative (<3), mBV-intermediate (>3-<7), or mBV-positive (7-10). Equivalent rulebased or machine-learned mappings may be used, provided they reproducibly partition samples into BV-negative / intermediate / positive states.

[0090] In some embodiments, the cervicovaginal microbiome profile further comprises a community state type (CST) assignment.

[0091] As used herein, the term “community state type (CST)” refers to a categorical classification of microbiome community composition. Commonly recognized states include those dominated by a single Lactobacillus species-CST-I (L. crispatus), CST-II (L. gasseri), CST-III (L. iners), and CST-V (L. jensenii)-and CST-IV, a polymicrobial state characterized by low Lactobacillus abundance and high diversity. CST-IV may be further subdivided; for example, CST-IV-A is characterized by a richness in anaerobes such as Gardnerella, Prevotella, and Candidatus Lachnocurva vaginae, while CST-IV-B is characterized by Atopobium. CST assignment may be performed using nearest-centroid classification, k-means clustering, or other trained classifiers.Docket No. 182219.00297

[0092] In some embodiments, CST is assigned by nearest-centroid classification using reference centroids (e.g., CST-I, II, III, V, and IV subtypes). Feature space may be CLR-transformed genus or species abundances; distance metrics may include Aitchison, Euclidean, or Bray-Curtis. Probability outputs can be generated by softmax over inverse distances, and a label is assigned to the maximum-probability class.

[0093] As used herein, the term “nearest-centroid classifier” refers to a supervised classification algorithm that computes distances between a sample feature vector and precomputed centroids for each class (e.g., CSTs) and assigns the class of the closest centroid under a specified distance metric.

[0094] As used herein, “an elevated level of Candidates Lachnocurva vaginae'" refers to an increased abundance of Candidatus Lachnocurva vaginae in a biological sample (e.g., vaginal fluid) relative to a reference or baseline level. Elevated abundance may be determined by a validated molecular assay, such as quantitative PCR, next-generation sequencing, or other nucleic acid-based techniques, and may be expressed as an absolute quantity (e.g., genome copies per mL) or as a relative proportion of total bacterial composition. In certain embodiments, an elevated level corresponds to a value above a predefined threshold established by clinical or algorithmic criteria, indicating a significant enrichment of this species compared to normal or healthy microbiota.

[0095] As used herein, “elevated” refers to an increase in the level, amount, or abundance of a specified microbial species, analyte, biomarker, or microorganism in a biological sample relative to a reference or baseline level. In some embodiments, “elevated” corresponds to an increase of at least about 1%, 2%, 5%, 10%, 20%, 25%, 50%, or more compared to the reference level. The reference level may be a normal or healthy control value, a pre-treatment value, or a population median, as appropriate for the context.

[0096] As used herein, the term “microbial risk score (MRS)” denotes a scalar score computed from abundances or log -ratios of BV-associated taxa relative to Lactobacillus taxa, optionally transformed by centered log-ratio (CLR) and weighted by effect sizes (e.g., odds ratios or logistic regression coefficients) trained to predict incident or recurrent infection. The MRS may be calibrated to output a probability using Platt scaling or isotonic regression and categorized by predetermined cutoffs.

[0097] As used herein, the term “BV-associated taxa” includes taxa empirically associated with BV or BV-like microbiomes, such as Candidates Lachnocurva vaginae (formerlyDocket No. 182219.00297

[0098] BVAB1), Gardnerella, Prevotella, Megasphaera, Atopobium, Sneathia, Finegoldia, Peptoniphilus, Anaerococcus, Clostridium, and others identified through differential abundance analysis and cross-validated model selection. The set may be refined for specific cohorts.

[0099] BV is one of the most common vaginal infections affecting women worldwide, BV results from an imbalance in the vaginal microbiota, generally in which the normally dominant Lactobacillus species are overrun by organisms such as Gardnerella vaginalis. BV can be classified into different subtypes based on the clades of bacteria that dominate the infection such as mBV-A and mBV-B. BV can be diagnosed using Amsel’s Criteria (Amsel R et al. Am J Med. 1983 Jan;74(l): 14-22) and Nugent Score (Nugent R. P. et al. Journal of Clinical Microbiology. 1991. 29:297-301) or molecular diagnostic techniques as known in the art (e.g., PCR test, etc.).

[0100] The vaginal microbiota is classified into five community state types (CSTs I-V) based on four Lactobacillus species. As used herein, the term “CST IV” can be classified into CST- IV-A, CST- IV-B, CST- IV-C0, CST- IV-C1,CST- IV-C2, CST- 1V-C3, and CST-IV-C4. CST-IV-A is characterized by the absence of Lactobacillus spp., but richness in Prevotella spp. and Peptoniphilus spp. CST-IV-B is characterized by low levels of lactobacillus spp., with low proportions of various anaerobic bacterial microorganisms and an abundance of Atopobium spp., Corynebacterium spp., Finegoldia spp., and Gardnerella spp. CST-IV-C1, CST-IV-C2, CST-IV- C3, CST-IV-and C4 are dominated by Prevotella spp., Enterococcus spp., Bifidobacterium spp,, and Staphylococcus spp., respectively.

[0101] An exemplary log-ratio R is defined as R = loglO[(abundance of Candidatus Lachnocurva vaginae + s) / (sum of Lactobacillus taxa + a)], with a set to a pseudo-count corresponding to 10-6relative abundance. In some embodiments, elevated risk is triggered when R exceeds a predetermined cutoff selected via ROC analysis (e.g., Youden’s J) or a percentile cutoff (e.g., >75th percentile) calibrated to achieve a target sensitivity (e.g., >80%).

[0102] As used herein, the term “CST-1V-A” refers to a cervicovaginal community state type sub-cluster enriched for anaerobes typical of BV, including but not limited to Candidatus Lachnocurva vaginae, Gardnerella, and Prevotella, and characterized by low Lactobacillus abundance.

[0103] As used herein, the term “incident infection” refers to a newly acquired infection not present at baseline and detected prospectively at follow-up (e.g., incident ChlamydiaDocket No. 182219.00297

[0104] trachomatis infection). As used herein, the term “reinfection” refers to a recurrent infection following treatment and apparent clearance.

[0105] As used herein, the term “mBV-A” refers to a subtype of molecular bacterial vaginosis (mBV) characterized by an mBV-positive state (molBV > 7 on a 0-10 scale) together with CST-IV-A and elevated abundance of Candidatus Lachnocurva vaginae relative to Lactobacillus, where elevation is determined by a predefined ratio cutoff, such as R = loglO[(Candidatus Lachnocurva vaginae + s) / (sum of Lactobacillus taxa + s)], selected by ROC analysis or fixed quantiles.

[0106] As used herein, the term “predetermined threshold” means a quantitative cutoff selected before classifying a subject. In various embodiments, the threshold is set by at least one of: receiver-operating-characteristic (ROC) analysis (e.g., Youden’s J optimization), fixed sensitivity or specificity targets (e.g., >80% sensitivity), calibration to an absolute risk target (e.g., probability >0.20 within 6-12 months), or anchoring to quantiles (e.g., top decile or quartile of score distribution) of a training cohort.

[0107] As used herein, the term “elevated risk” means a classification that corresponds to an increased likelihood of incident or recurrent disease relative to a reference group. In some embodiments, elevated risk corresponds to an odds ratio of at least 1.5, at least 2.0, or at least 3.0 versus the baseline category, or to a predicted probability at or above 0.15, 0.20, 0.25, or 0.30 within a prespecified time horizon (e.g., 6-12 months). In some embodiments, “elevated risk” denotes a classification category indicating increased likelihood of incident or recurrent infection or adverse outcome relative to a baseline or reference population, based on predefined thresholds of MRS, mBV category, and / or CST subtype, optionally conditioned on covariates.

[0108] In some embodiments, exemplary quantitative thresholds are used for risk classification. For instance, an mBV-positive state can correspond to a molBV score > 7 on a 0-10 scale. An elevated risk from the ratio R can be indicated when R is above a cohort- calibrated cutoff derived via Youden’s J from ROC analysis. High risk can be denoted by an MRS probability > 0.20 for an adverse outcome within a 6-12 month period.

[0109] In some embodiments, the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis;Docket No. 182219.00297

[0110] urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

[0111] To support applicability across the above-listed diseases or disorders, the dysbiosis markers (e.g., mBV-A, MRS) are linked to plausible mechanistic pathways such as inflammation and mucosal barrier disruption relevant to each condition. The computational pipeline may be re-calibrated using disease-specific cohorts to derive indication-specific risk thresholds, providing a unified yet adaptable framework for risk assessment.

[0112] As used herein, the term “therapy” includes microbiome-targeted interventions (e.g., Lactobacillus probiotics, prebiotics, vaginal microbiota transplant), non-antibiotic biofilm-disrupting agents (e.g., lactic acid gels, boric acid suppositories, disulfide bond disruptors), antibiotic stewardship strategies, or behavioral counseling, administered to reduce risk.

[0113] In some embodiments, the therapy comprising a microbiome-targeted therapy that comprises vaginal Lactobacillus probiotic therapy and / or a non-antibiotic biofilm-disrupting agent. In some embodiments, when elevated risk is identified, the following non-limiting microbiome-targeted interventions may be administered alone or in combination, under clinician supervision.

[0114] In some embodiments, the therapy comprises vaginal Lactobacillus probiotic formulations comprising one or more of L. crispatus, L. jensenii, L. gasseri, or L. rhamnosus at per-dose amounts between about 107and 1010CFU, administered intravaginally daily for about 5—14 days followed by weekly maintenance dosing for about 4-12 weeks.

[0115] In some embodiments, the therapy comprises boric acid intravaginal suppositories (e.g., about 300-600 mg) once daily for about 7-14 days, with optional maintenance dosing (e.g., 2- 3 times per week) for about 4-12 weeks, particularly in the setting of recurrent dysbiosis.

[0116] In some embodiments, the therapy comprises lactic-acid-containing gels or other non-antibiotic biofilm-disrupting agents applied intravaginally per product instructions on a schedule such as once daily for an induction period (e.g., 5-10 days) then 1-3 times per week for maintenance.Docket No. 182219.00297

[0117] In some embodiments, dosing, duration, and agent selection can be individualized based on tolerance, concomitant conditions, and local standards of care. Antibiotic stewardship strategies can include deferring antibiotics when not indicated and prioritizing microbiome-supportive therapies where appropriate.

[0118] In some embodiments, the method comprises computing the microbial risk score by: performing amplicon sequence variant inference on 16S reads; assigning taxonomy with a trained classifier; transforming taxa abundances via centered log-ratio; computing log-ratios relative to Lactobacillus; and applying pre-trained weights with Platt-scaled calibration to output a probability of incident infection.

[0119] In some embodiments, the microbiome profile further comprises a ratio of Candidatus Lachnocurva vaginae to Lactobacillus, and the subject is classified as elevated risk when the ratio exceeds a predetermined cutoff.

[0120] In some embodiments, the plurality of BV-associated taxa used to compute the microbial risk score comprises at least Candidatus Lachnocurva vaginae and one or more of Prevotella, Megasphaera, Clostridium, Staphylococcus, Acinetobacter, or additional taxa selected by differential abundance analysis and cross-validation.

[0121] In some embodiments, the method further comprises outputting a report comprising the subject’s risk classification, the mBV category, and the microbial risk score value, and optionally a CST assignment.

[0122] In some embodiments, the risk assessment is prospective, based on a sample collected prior to detection of incident infection.

[0123] In some embodiments, the method assesses risk of reinfection following antibiotic treatment, based on a post-treatment sample.

[0124] In some embodiments, the molBV score is computed by transforming 16S rRNA gene amplicon sequence features into a Nugent-like score and classifying into negative, intermediate, or positive categories. In some embodiments, the molBV score is derived by mapping ASV or genus-level features onto a Nugent-like index. In one implementation, features corresponding to Lactobacillus morphotypes and BV-associated anaerobes are combined into an index scaled from 1 to 10, with cutoffs set to categorize “negative,” “intermediate,” and “positive ” For example, high relative abundance of Lactobacillus crispatus and / or L. jensenii yields low scores, while elevated abundances of Gardnerella,Docket No. 182219.00297

[0125] Candidatus Lachnocurva vaginae, Prevotella, Megasphaera, and Atopobium drive higher scores. The mapping can be rule-based or learned via regression trained to predict Nugent categories from paired Gram-stain references. The model is calibrated to preserve clinical operating points (e.g., sensitivity / specificity for BV status).

[0126] In some embodiments, the microbial risk score is a weighted sum of taxa abundances or log-ratios relative to Lactobacillus and the weights are derived from odds ratios for incident infection. In some embodiments, the microbial risk score is scaled and categorized using predetermined cutoffs corresponding to increasing odds of incident infection.

[0127] In some embodiments, the method comprises: (1) fitting a multivariable logistic regression predicting incident infection at tO using features from t- 1; obtain coefficients βi; (b) computing MRS_raw = Σi βi·xi, where xi are feature values; (c) calibrating MRS raw to predicted probability p via Platt scaling (sigmoid) or isotonic regression on a validation set, yielding MRS_prob in [0,1]; and defining risk categories by predetermined thresholds, e.g., Low (<0.10), Intermediate (0.10-<0.20), High (>0.20), or by cohort-specific ROC-optimized cut points.

[0128] In some embodiments, the method further comprises integrating a sexual risk behavior score and / or high-risk HPV status as covariates into the risk classification model.

[0129] In some embodiments, the classification step is performed by a trained machine-learning model that outputs a categorical risk label and / or a probability score. In some embodiments, the classifier is calibrated such that mBV-in termediate yields an intermediate risk category and mBV-A yields a higher risk category.

[0130] In some embodiments, the method further comprises recommending an intervention selected from follow-up testing, antibiotic stewardship, microbiome-targeted therapy, prebiotic or probiotic therapy, or behavioral counseling based on the elevated risk classification. Risk assessment is prospective when computed from a sample collected prior to detection of incident infection. The pipeline also supports evaluation of reinfection risk after antibiotic treatment, using post-treatment samples to assess persistence or emergence of mBV-A. Longitudinal reports compare pre-infection and post-treatment states, flagging transitions toward mBV- positive or increasing Candidatus Lachnocurva vaginae abundance as increased risk trajectories and recommending surveillance intervals accordingly.Docket No. 182219.00297

[0131] Elevated risk classification can trigger recommendations, including follow-up testing, microbiome-targeted therapy (e.g,, vaginal Lactobacillus probiotics), non-antibiotic biofilm-disrupting agents, prebiotic supplementation, behavioral counseling, or antibiotic stewardship practices. In treatment claims, when elevated risk is identified, a therapy is administered according to a dosing regimen effective to reduce risk. Example regimens include daily or cyclical intravaginal Lactobacillus crispatus probiotic for 4-12 weeks, optionally followed by maintenance dosing; or a course of biofilm-disrupting gel nightly for 5—10 days followed by probiotics.

[0132] In some embodiments, the sample is a cervical cytology specimen collected in a liquid-based cytology medium. In some embodiments, the analysis is performed on archived DNA from a Pap-smear specimen.

[0133] In some embodiments, the method further comprises detecting high-risk HPV genotypes and reporting the combined microbiome and HPV risk profile.

[0134] In some embodiments, the mBV-A subtype is defined by an mBV-positive state combined with elevated abundance of Candidatus Lachnocurva vaginae, and the subject is classified as elevated risk upon identification of mBV-A.

[0135] In some embodiments, the elevated risk classification triggers a programmed testing schedule with an increased surveillance interval.

[0136] In some embodiments, the method outputs a longitudinal comparison between a pre-infection sample and a post-treatment sample, highlighting persistence or emergence of the mBV-A subtype.

[0137] In some embodiments, the pipeline flags subjects exhibiting a transition toward mBV- positive states as at risk of incident infection.

[0138] In another aspect, this disclosure provides a method of guiding clinical management for a disease or disorder. In some embodiments, the method comprises: performing the method as described herein to identify a subject at elevated risk of the disease or disorder; and recommending at least one of microbiome-targeted therapy, probiotic therapy, or adjusted follow-up intervals.

[0139] Systems for Diagnosing and Treating Diseases or DisordersDocket No. 182219.00297

[0140] In yet another aspect, this disclosure further provides a computer-implemented system for assessing risk of a disease or disorder in a subject with a vagina and / or cervix. In some embodiments, the system comprises: (a) a data processor and memory storing instructions that, when executed, cause the system to: receive subject-level microbiome data comprising bacterial 16S rRNA gene sequencing features from a vaginal or cervicovaginal sample; compute a molBV score and optionally assign a CST; and compute a microbial risk score from a set of BV-associated taxa; and classify a risk of Chlamydia trachomatis infection in the subject based on: (i) molBV-positive with an elevated level of Candidatus Lachnocurva vaginae, or (ii) the microbial risk score exceeding a threshold; and (b) an interface configured to generate a report comprising the risk classification.

[0141] The in vitro data processing and classification methods described herein generate scores, labels, and reports for clinician consideration and do not, by themselves, establish a medical diagnosis.

[0142] As used herein, the term “processor” refers to any hardware device capable of executi ng program instructions, including central processing units (CPUs), graphics processing units (GPUs), tensor processing units (TPUs), system-on-chip (SoC) devices, field-programmable gate arrays (FPGAs), or combinations thereof.

[0143] As used herein, the term “threshold” refers to a pre-specified or adaptively learned decision boundary used to dichotomize or stratify continuous scores (e g., MRS, molBV) into risk categories, including static thresholds, percentile-based thresholds, or thresholds updated through model training procedures.

[0144] In some embodiments, the system implements data preprocessing steps comprising sequence quality filtering, denoising, chimera removal, taxonomic assignment using a reference classifier, and normalization of feature tables to relative abundance or compositional transformations, such as centered log-ratio transformation, with zero-handling procedures. The system may compute ratio metrics including a ratio of Candidatus Lachnocurva vaginae to Lactobacillus spp. and incorporate such ratio metrics into the risk classification. The system may compute diversity metrics, including alpha and beta diversity, and derive composite features aggregating BV-associated taxa. Feature sets may be selected using univariate effect sizes or multivariate regularization methods.

[0145] In some embodiments, the microbial risk score is calculated as a weighted sum of abundances or log-ratios of taxa relative to Lactobacillus, with weights derived from effectDocket No. 182219.00297

[0146] sizes for incident infection. Decision logic may classify a subject at elevated risk of Chlamydia trachomatis infection if the subject is mBV-positive with an elevated level of Candidatus Lachnocurva vaginae or if the MRS exceeds a threshold. In alternative embodiments, the classifier is a logistic regression, gradient boosting machine, random forest, support vector machine, or neural network operating on the feature vector comprising microbiome features and optional covariates. Thresholds may be determined by maximizing a performance criterion, such as the Youden index. Fl score, or a predefined sensitivity target, and may be recalibrated through post-training calibration methods.

[0147] As used herein, the term “ratio metrics” refers to features computed as ratios or logratios between abundances or transformed abundances of two or more taxa or groups of taxa, including but not limited to a ratio of Candidatus Lachnocurva vaginae to Lactobacillus spp.

[0148] In some embodiments, the system can expose a user interface or application programming interface for data upload and returns outputs comprising a risk label, probability or risk score, mBV category, optional CST assignment, MRS value, ratio metrics, time stamps, quality flags including sequencing depth and target read count, and recommended next steps consistent with the specification. Reports may include interpretability annotations, feature importance summaries, and disclaimers for research use or clinical decision support, as applicable by jurisdiction.

[0149] In some embodiments, a computer-implemented system comprises at least one processor, system memory, non-transitory storage, and a network interface. Instructions, when executed, perform: (a) ingest of sample-level features (e.g., ASV tables, taxonomy, metadata); (b) computation of molBV, mBV, CST, and MRS as described; (c) optional integration of covariates such as sexual risk behavior score and high-risk HPV status via concatenation into the feature vector or as separate inputs to a joint model; (d) classification into risk categories and generation of a human-readable report.

[0150] In some embodiments, the system further comprises a model-training module configured to update classification thresholds or weights based on newly ingested labeled data. The module may implement standard cross-validation schemes, hold-out validation for final performance estimation, hyperparameter optimization, and drift monitoring. Drift monitoring may track changes in input feature distributions, target prevalence, and calibration stability, triggering alerts or retraining workflows. Model artifacts may be persisted as serialized objectsDocket No. 182219.00297

[0151] with versioning and metadata capturing data schema, preprocessing steps, and performance metrics.

[0152] As used herein, the term “drift monitoring” refers to procedures for detecting changes over time in data distributions, model inputs, model outputs, or performance metrics that may degrade model generalization, including data drift, concept drift, and covariate shift.

[0153] As used herein, the term “training data” refers to data with associated labels or outcomes used to fit model parameters, including incident or prevalent disease outcomes, and metadata such as demographic, behavioral, and clinical covariates.

[0154] In some embodiments, implementations may be in Python, R, Java, C / C++, or equivalents; models may be persisted as serialized objects (e.g., pickled or ONNX) with versioning. The system can expose a user interface or API for data upload and returns outputs comprising: risk label, probability, mBV category, CST assignment, MRS value, ratio metrics, time stamps, quality flags (e.g., sequencing depth), and recommended next steps consistent with the specification. A model-training module can update weights or thresholds upon ingestion of labeled data, using standard cross-validation, hold-out validation, and drift monitoring.

[0155] In some embodiments, CST assignment is performed using a nearest-centroid classifier trained on reference community profiles, hierarchical clustering with silhouette-based model selection, or a supervised learner mapping compositional features to CST labels. In some embodiments, molBV is computed using a supervised model trained to predict Nugent score or Amsel diagnosis from sequencing features, optionally incorporating qPCR-derived abundance priors. In some embodiments, the MRS is constructed as a sum of weighted logratios of BV-associated taxa over Lactobacillus reference groups, with regularization to prevent overfitting and with stability selection to ensure reproducibility. In some embodiments, classification outputs include calibrated probabilities with uncertainty intervals derived from bootstrapping or Bayesian posteriors. In some embodiments, the user interface validates input schema, provides data provenance and processing summaries, and flags samples not meeting quality thresholds. In some embodiments, the system supports batch processing and parallelization to handle high-throughput sequencing datasets, with job orchestration and resource allocation telemetry.

[0156] In some embodiments, the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; MycoplasmaDocket No. 182219.00297

[0157] genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

[0158] In some embodiments, the instructions further cause the system to compute a ratio of Candidatus Lachnocurva vaginae to Lactobacillus and incorporate the ratio into the risk classification. The ratio may be expressed as a relative abundance ratio or as a log-ratio under a compositional data framework and weighted in the MRS or used as an independent decision rule.

[0159] In some embodiments, the microbial risk score is calculated as a weighted sum of abundances or log-ratios of taxa relative to Lactobacillus, with weights derived from effect sizes for incident infection.

[0160] In some embodiments, the system further comprises a model-training module configured to update classification thresholds or weights based on newly ingested labeled data. Updating may be performed on a schedule or triggered by drift monitoring, with safeguards including validation on a hold-out set and preservation of prior model versions for rollback.

[0161] In another aspect, this disclosure provides a non-transitory computer-readable medium storing instructions that, when executed by a processor, cause a computing device to perform the method as described herein to identify a subject at elevated risk of the disease or disorder. The instructions may include modules for data ingestion, preprocessing, feature computation including molBV, mBV, CST, and MRS, classification and calibration, report generation, and optional model training and drift monitoring.

[0162] As used herein, the term “non-transitory computer-readable medium” refers to any tangible storage medium that is not a transitory propagating signal, including magnetic, optical, and semiconductor memory devices, storing instructions executable by a processor.

[0163] Kits for Diagnosing and Treating Diseases or DisordersDocket No. 182219.00297

[0164] In another aspect, this disclosure additionally provides a kit for assessing risk of a disease or disorder in a subject with a vagina and / or cervix. In some embodiments, the kit comprises: (a) reagents for extraction of nucleic acids from a vaginal / cervicovaginal sample; (b) a primer set configured to amplify bacterial 16S rRNA gene sequences from the sample; (c) optionally, a primer set configured to amplify fungal or eukaryotic ITS sequences from the sample; (d) one or more positive controls and negative controls for sequencing and analysis quality assurance; and (e) instructions to access software or an analysis pipeline configured to compute an molBV score, optionally assign a CST, compute a microbial risk score from BV-associated taxa, and classify infection risk based on: (i) mBV-positive with an elevated level of Candidatus Lachnocurva vaginae, or (ii) the microbial risk score.

[0165] As used herein, the term “nucleic acid” refers to DNA and / or RNA, including genomic DNA, plasmid DNA, mitochondrial DNA, ribosomal RNA, messenger RNA, and complementary DNA, as well as fragments, amplicons, adapters, and barcodes derived therefrom or appended thereto.

[0166] As used herein, the term “reagent for extraction of nucleic acids” refers to any buffer, enzyme, chaotrope, detergent, salt, organic solvent, bead, column, membrane, magnetic particle, or consumable that facilitates lysis, protein and inhibitor removal, nucleic acid binding, washing, elution, stabilization, or storage of nucleic acids from the sample.

[0167] As used herein, the term “primer” or “primer set” refers to one or more oligonucleotides configured to anneal to a target region and enable amplification by polymerase, including barcoded, indexed, adapter-tailed, or fusion primers suitable for library preparation for sequencing.

[0168] As used herein, the term “barcoded” refers to inclusion of single or dual index sequences, unique molecular identifiers (UMIs), or inline sample identifiers that enable multiplexing, demultiplexing, error correction, deduplication, or assignment of reads to samples or controls.

[0169] As used herein, the term “16S rRNA gene” refers to the bacterial small subunit ribosomal RNA gene and its hypervariable regions, including but not limited to V1-V2, V3- V4, V4, and V4-V5, or V5-V7, that permit taxonomic discrimination among bacteria.Docket No. 182219.00297

[0170] As used herein, the term “ITS” or “internal transcribed spacer” refers to the fungal or eukaryotic internal transcribed spacer regions (e.g., ITS1 and / or ITS2) used for taxonomic profiling of fungi or other eukaryotic microorganisms.

[0171] As used herein, the term “amplicon sequence variant” or “ASV” refers to inferred biological sequences resolved by denoising algorithms from amplicon data at single-nucleotide resolution, as distinguished from operational taxonomic units clustered at fixed similarity' thresholds.

[0172] In some embodiments, a kit comprises nucleic acid extraction reagents sufficient for at least 20, 50, or 100 samples, with buffer concentrates and wash solutions stabilized for storage at ambient or refrigerated temperatures. In certain embodiments, lysis buffers include chaotropic salts and detergents compatible with silica-based or magnetic bead-based binding, and eluates are compatible with downstream PCR and library preparation without additional cleanup.

[0173] In some embodiments, a kit comprises a barcoded primer set targeting bacterial 16S (e.g., V4 or V3-V4 regions) and optionally a barcoded ITS primer set. Each primer may be provided at 10-100 pM in nuclease-free water or buffer, optionally in single-use aliquots to minimize freeze-thaw cycles. In certain embodiments, primers include platform-specific adapter tails, sample indices, and sequencing-compatible spacers to minimize low-diversity issues.

[0174] In some embodiments, a kit comprises a positive control, such as a defined mock community including Lactobacillus and BV-associated taxa at known relative abundances; and a negative control comprising nuclease-free water. In some embodiments, the positive control comprises a defined mock community including Lactobacillus crispatus, Lactobacillus iners, Candidatus Lachnocurva vaginae, and Prevotella bivia, optionally stabilized in a lyophilized format for ambient shipment and reconstitution.

[0175] In some embodiments, a kit comprises instructions to access software or a pipeline implementing molBV, mBV, CST, and MRS computations with versioned reference files and example datasets. The instructions can include secure access details, required input formats, compatible sequencing platforms, and version identifiers for classifiers and reference databases.Docket No. 182219.00297

[0176] In some embodiments, a kit comprises optional sample collection devices, such as flocked swabs, brushes, or self-collection swabs, and transport media validated to preserve microbial DNA for at least 7-28 days at 2-8°C and at least 24-72 hours at ambient temperature. In certain embodiments, transport media are formulated to inactivate nucleases, stabilize cell-free DNA, and reduce overgrowth. In some embodiments, the kit further comprises a sample collection device and transport medium suitable for preserving vaginal or cervicovaginal specimens for microbiome analysis, with lot-specific certificates of analysis indicating sterility, nucleic acid background levels, and validated preservation durations.

[0177] In some embodiments, the written instructions can specify acceptance criteria, such as minimum read depth, positive / negative control recovery windows, barcode collision thresholds, and maximum contaminant read proportions; data upload format, such as ASV table and taxonomy file; and interpretation rules, such as elevated risk upon mBV-positive with CST-IV-A or MRS above a threshold. In further embodiments, the instructions specify corrective actions upon QC failure and criteria for re-extraction or resequencing.

[0178] In some embodiments, the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

[0179] In some embodiments, the software implements a nearest-centroid classifier to assign CSTs and a trained risk model that outputs a risk label or probability for incident or recurrent infection. In some embodiments, the software reports a ratio of Candidatus Lachnocurva vaginae to Lactobacillus and compares the ratio to a predetermined cutoff associated with elevated risk. In some embodiments, the software enables integration of covariates comprising sexual risk behavior metrics and high-risk HPV status into the risk classification and reports both microbiome-only and integrated risk outputs. In some embodiments, the software allows user configuration of thresholds for elevated risk, while also providing default thresholdsDocket No. 182219.00297

[0180] derived from validation studies, and includes safeguards to prevent misclassification due to insufficient data or QC failure. In further embodiments, risk outputs are accompanied by interpretive comments recommending confirmatory diagnostic testing, clinical evaluation, or follow-up sampling intervals.

[0181] In some embodiments, the kit further comprises written instructions indicating that elevated risk is determined upon identification of an mBV-positive state with an elevated level of Candidatus Lachnocurva vaginae or a microbial risk score above a threshold, and recommending confirmatory testing or clinical follow-up.

[0182] As used herein, the term “positive control” refers to a sample of known composition, titer, or expected result included to verify assay performance, sequencing, and analysis accuracy. “Negative control” refers to a no-template or background control, such as nuclease- free water, used to detect contamination or spurious amplification.

[0183] In some embodiments, the positive control comprises a defined mock community including Lactobacillus and selected BV-associated taxa, and the negative control comprises nuclease-free water. In some embodiments, the positive control comprises a defined mock community including Lactobacillus crispatus, Lactobacillus iners, Candidatus Lachnocurva vaginae, and Prevotella bivia

[0184] As used herein, the term “analysis pipeline” refers to a set of computational steps, scripts, software modules, and reference files that process raw sequencing reads through quality control, denoising, chimera removal, taxonomic assignment, feature table generation, transformation, and computation of scores, classifications, and risk outputs. In some embodiments, the analysis pipeline computes the molBV score by transforming 16S data into a Nugent-like score on a 0-10 scale and classifies the score into negative, intermediate, or positive categories. In some embodiments, the pipeline produces both prospective risk for incident infection and risk of reinfection following treatment, based on pre- and post-treatment sampling.

[0185] In some embodiments, the analysis pipeline stores and reports run metadata, including instrument identifiers, run dates, read lengths, trimming parameters, classifier versions, and database versions, and generates an audit trail. In certain embodiments, outputs include mBV category, CST assignment with confidence scores, microbial risk score with threshold interpretation, and integrated risk probability with 95% confidence intervals or credible intervals when applicable.Docket No. 182219.00297

[0186] In some embodiments, raw reads undergo adapter trimming, quality filtering, and denoising to infer amplicon sequence variants (ASVs). DADA2 or similar ASV inference methods may be used. Chimeric sequences are removed. Taxonomy is assigned using a trained classifier (e.g., naive Bayes trained on a curated 16S database such as SILVA or Greengenes with region-specific trimming). Ambiguities may be resolved at genus level where species¬ level resolution is not reliable. Read count tables are transformed to relative abundances. To address compositionality, centered log-ratio (CLR) transformation is optionally applied after adding a small pseudocount.

[0187] In some embodiments, the microbial risk score is derived from a validated set of BV-associated taxa including at least Candidatus Lachnocurva vaginae and Prevotella, optionally including Gardnerella, Atopobium vaginae, Sneathia, Megasphaera, and Mobiluncus, with weights established by logistic regression, gradient boosting, or other supervised learning methods.

[0188] In some embodiments, the primer set targets the V1-V3, V3-V4, V4-V5, or V5-V7 region of the bacterial 16S rRNA gene. In other embodiments, the primer set targets the V4 region to maximize compatibility with existing reference classifiers. Primer sequences may be optimized to minimize host DNA amplification and to balance coverage across Lactobacillus and BV-associated taxa.

[0189] In some embodiments, the kit includes a container with the primers and optionally, informational material. The informational material can be descriptive, instructional, marketing, or other material related to the methods described herein. The informational material of the kits is not limited in its form. In some embodiments, the informational material can include information about production of primers, concentration, batch or production site information, and quality control data. In one embodiment, the informational material relates to methods of using the primers in a PCR reaction. The information can be provided in various formats, including printed text, computer-readable material, video recording, audio recording, or information that contains a link or address to substantive material.

[0190] In some embodiments, the container is a bottle or vial and the informational material can be contained in a plastic sleeve or packet. In some embodiments, the kit components are provided with unique lot identifiers and expiration dating aligned to validated stability studies at ambient and refrigerated temperatures. In some embodiments, the extraction reagents includeDocket No. 182219.00297

[0191] carrier RNA to improve nucleic acid recovery from low biomass samples, and the kit provides low-retention tubes certified DNA / RNA-free.

[0192] Additional Definitions

[0193] As used herein, the phrases “in one embodiment,” “in various embodiments,” “in some embodiments,” and the like are used repeatedly. Such phrases do not necessarily refer to the same embodiment, but they may unless the context dictates otherwise.

[0194] As used herein, the terms “and / or” or “ / ” mean any one of the items, any combination of the items, or all of the items with which this term is associated.

[0195] As used herein and in the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise.

[0196] As used herein, the term “approximately” or “about,” as applied to one or more values of interest, refers to a value that is similar to a stated reference value. In some embodiments, the term “approximately” or “about” refers to a range of values that fall within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less in either direction (greater than or less than) of the stated reference value unless otherwise stated or otherwise evident from the context (except where such number would exceed 100% of a possible value). Unless indicated otherwise herein, the term “about” is intended to include values, e.g., weight percents, proximate to the recited range that are equivalent in terms of the functionality of the individual ingredient, the composition, or the embodiment.

[0197] As used herein, the term “each,” when used in reference to a collection of items, is intended to identify an individual item in the collection but does not necessarily refer to every item in the collection. Exceptions can occur if explicit disclosure or context clearly dictates otherwise.

[0198] As disclosed herein, a number of ranges of values are provided. It is understood that each intervening value, to the tenth of the unit of the lower limit, unless the context clearly dictates otherwise, between the upper and lower limits of that range is also specifically disclosed. Each smaller range between any stated value or intervening value in a stated range and any other stated or intervening value in that stated range is encompassed within the present disclosure. The upper and lower limits of these smaller ranges may independently be included or excluded in the range, and each range where either, neither, or both limits are included inDocket No. 182219.00297

[0199] the smaller ranges is also encompassed within the present disclosure, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the present disclosure.

[0200] The following examples serve to further illustrate the methods of the present disclosure.

[0201] Examples

[0202] Example 1. Materials and methods

[0203] This Example describes the materials and methods used in Example 2.

[0204] STAR Methods

[0205] Experimental Model And Study Participant Details

[0206] Participants were selected from a longitudinal dynamic cohort study of sexually active adolescent female patients receiving gynecological care at Mount Sinai Adolescent Health Center (MSAHC) in New York City. Overall population characteristics, eligibility criteria (which included female sex at birth and history of sexual intercourse), and study design of the larger cohort have been previously described. (Schlecht, N. F. et al. (2012) PLoS One 7, 37419.). Briefly, study participants were enrolled between the ages of 13-21 and received a gynecological examination at each study visit, approximately every 6 months. Routine screening for gonorrhea and chlamydia was performed at each visit and as clinically indicated using GEN-PROBE APTIMA (Hologic, Marlborough, MA) assays performed by the clinical laboratory. Clinical indications for testing included patient reports of a positive exposure (i.e., partner with a known STI), or a patient’s own symptoms (e.g., vaginal discharge, itch, pelvic pain, and / or bleeding with intercourse). In addition, patients could request to test whether the CT infection was resolved, which was usually done within 4 weeks after treatment. Research questionnaires were administered at every study visit, which included questions on age, sex, race, ethnicity, place of birth, sexual behaviors, sexual partners, history of STIs, condom use, drug and alcohol use, and other related factors. (Braun-Courville, D. K. et al. (2014) J. Pediatr. Adolesc. Gynecol. 27, el03-el08). All participants were of female sex per eligibility criteria for the study, and gender identity was not collected for analysis purposes. Written informed consent was obtained from all participants prior to enrollment. The Institutional Review Board at Icahn School of Medicine at Mount Sinai and Mount Sinai Hospitals Group approved the study (Federalwide Assurance # FWA00005651).Docket No. 182219.00297

[0207] Key Resource Table.

[0208] REAGENT or RESOURCE SOURCE IDENTIFIER Chemicals, peptides, and recombinant proteins

[0209] QIAamp Mini spin column Qiagen, Valencia, CA 51306

[0210] Platinum 10X PCR buffer Invitrogen, Waltham, MA 10968018

[0211] MgCl2Applied Biosystems, Carlsbad, 4311806

[0212] CA

[0213] dNTP mix Roche, Basel, Switzerland 11814382001 AmpliTaq Gold® Polymerase Applied Biosystems. Carlsbad, 4311806

[0214] CA

[0215] Platinum™ Taq DNA Invitrogen, Waltham, MA 10968018

[0216] Polymerase

[0217] Deposited data

[0218] 16SV4, rRNA & ITS1 Qiita Qiita Repository Study ID: sequencing reads 14884S

[0219] Software and algorithms

[0220] molBV GitHub github.com / musyk07 / molBV VALENCIA GitHub gi thub. com / ravel -

[0221]

[0222] lab VALENCIA

[0223] Nested case-control study design

[0224] A nested case-control study design was used to select cases with incident CT detected after enrollment into the cohort, i.e., CT negative at a ti and positive at to visit. Participant samples were selected as part of a prospective open cohort, which began enrollment in October 2007 and continued through this analysis using all data until October 2022. Thus, enrollment and follow-up occurred during this interval. Two control individuals (CT negative at ti and to) were then matched to each case on age (±3 years), year of enrollment, and prior history of CT. Controls were allowed to become cases if they became CT positive later in the study.

[0225] Banked DNA samples for analysis of the CVM were obtained from specimens collected at the cohort study visits using a cervical Cytobrush® placed in PreservCyt™ transport medium (ThinPrep®; Hologic) following the same procedure as for Pap smears. Up to three vaginal or cervicovaginal samples from each participant were tested for the CVM (see FIG. 1 for study design). In a nested case-control design, cases can act as controls and vice versa (e.g., at the ti and t+i visits). Additionally, the participants are matched on person-time, come from the same population, which helps reduce selection bias and improve the comparability between groups.

[0226] A total of 187 participants were identified with an incident CT infection during the cohort study follow-up at the time of selection (January 2020) and were matched to 373 controls (one case had only a single matched control available for inclusion). Cervicovaginal samplesDocket No. 182219.00297

[0227] collected prior to the infection (t-i) were available for all 187 cases with a median (IQR) time prior to incident CT detection (to) of 6.75 (3.35) months, with a comparable sample available from 373 controls. Samples tested at the visit after CT detection and matched visit for controls were collected 6.70 (1.97) and 6.83 (2.83) months after to for cases and controls, respectively.

[0228] Method Details

[0229] Vaginal or cervicovaginal samples were stored immediately at -20°C until transport to the research lab at the Albert Einstein College of Medicine. The samples were transferred to a 15 ml tube in the lab and gently centrifuged at 1500 RPM for 5 min. After removing the supernatant by decanting, the pellets were rinsed in 3 ml of TE (10 mM Tris, 1.0 mM EDTA). This solution was then vortexed and centrifuged at 1500 RPM for 5 min and the supernatant was removed by decanting. The remaining pellet and leftover solution (—150 ml) were used for DNA isolation via column processing with the QIAamp Mini spin column (Qiagen, Valencia, CA) following the manufacturer’s protocol. The purified DNA was eluted in 150 ml of elution buffer (1 OmM Tris / 0.5mM EDTA, pH 9) and used initially for HPV DNA analyses and then stored at -20°C in a non-frost-free freezer. PCR for bacterial communities was performed using forward (515F) GTGYCAGCMGCCGCGGTA (SEQ ID NO: 1) and reverse (806R) GGACTACHVGGGTWTCTAAT (SEQ ID NO: 2) primers that amplify the V4 hypervariable region of the prokaryotic 16S rRNA gene. (Caporaso, J. G. et al. (2012) ISME J. 6, 1621-1624; Wang, Y., and Qian, P. Y. (2009) PLoS One 4, e7401). All primers contained unique Golay barcodes to allow for dual indexing of each sample and were purchased from IDT (IDT, Coralville, IA). PCRs were conducted in a 25 ml reaction with 2 ml input of template DNA, 16.75 ml of ddH2O, 2.5 ml of Platinum 1 OX PCR buffer (Invitrogen, Waltham, MA), 0.75 ml of MgCh (50 mM, Invitrogen), 0.5 ml of dNTP mix (lOmM each, Roche, Basel, Switzerland), 0.25 ml AmpliTaq Gold®, polymerase (5 U / ml, Applied Biosystems, Carlsbad, CA), 0.25 ml of Platinum® Taq DNA Polymerase (10 U / ml, Invitrogen), and 1 ml (5 mM) of each primer. Thermocycling conditions included an initial denaturation at 95°C for 5 min, followed by 15 cycles of 95°C for 1 m, 55°C for 1 m, 72°C for 1 m, followed by 15 cycles of 95°C for 1 m, 60°C for 1 m, 72°C for 1 m, and a final extension at 72°C for 10 min.

[0230] In order to more comprehensively profile the CVM, ITS1 sequencing was performed, which identifies fungal and eukaryotic species. These organisms have been reported to modulate the CVM despite their relatively minor biomass as compared to bacteria. (Bradford, L. L., and Ravel, J. (2017) Virulence 8, 342-351). PCR for eukaryotic communities was performed usingDocket No. 182219.00297

[0231] barcoded forward (48F) ACACACCGCCCGTCGCTACT (SEQ ID NO: 3) and reverse (217R) TTTCGCTGCGTTCTTCATCG (SEQ ID NO: 4) primers (IDT) that amplify the ITS1 region of the prokaryotic ribosomal gene cluster. (Usyk, M. et al. (2017). Novel ITS1 Fungal Primers for Characterization of the Mycobiome. mSphere 2, e00488-17; Rosenbaum, J et al. (2019) Sci. Rep. 9, 1531). PCRs were conducted in a 25 ml reaction with 10 ml input of template DNA, 8.75 ml of ddH20, 2.5 ml of Platinum lOx PCR buffer (Invitrogen), 0.75 nil of MgCh (50 mM, Invitrogen), 0.5 ml of dNTP mix (10 mM each, Roche), 0.25 ml AmpliTaq Gold® polymerase (5 U / ml, Applied Biosystems), 0.25 ml of Platinum® Taq DNA Polymerase (10 U / ml, Invitrogen), and 1 ml (5 mM) of each primer (IDT, Coralville, IA), Thermocycling conditions included an initial denaturation at 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 55°C for 30 s, 72°C for 2 min, followed by a final extension at 72°C for 10 min. All PCRs were conducted in a Veriti™ Thermal Cycler (Applied Biosystems, Foster City, CA) and PCR products were verified by gel electrophoresis.

[0232] PCR products for each sample were pooled by PCR assay (16S and ITS1 ) in approximately equal concentrations and 100 nil of the pooled products were loaded into a 3% agarose gel and run at 80V for 3 h to separate the DNA fragments. The DNA fragment for each assay was excised and purified with a QIAquick® Gel Extraction Kit (Qiagen) and quantified using a Qubit High Sensitivity dsDNA assay (Invitrogen). NGS library preparation was conducted on the purified pooled PCR products from each assay with a KAPA LTP Library Preparation Kit (KAPA Biosystems, Wilmington, MA) according to the manufacturer’s protocol. The library amplicons were validated on a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA) and sequencing of libraries was carried out on an Illumina NovaSeq 6000 using the 2x250 bp paired-end reads kit.

[0233] Quantification And Statistical Analysis

[0234] Sequence reads were clustered into amplicon sequence variants (ASVs) using DADA2 and taxonomy was assigned using a custom cervicovaginal microbiome specific database (Usyk, M. et al. (2022) Nat. Commun. 13, 233) employing a Naive Bayesian classifier. The presence of bacterial vaginosis (BV) was assessed using a previously validated molecular score from the 16SV4 reads (molBV). The molBV algorithm converts 16S rRNA gene amplicon sequences into a Nugent-like molBV score from 0-10. This assay provides a measure of the cervicovaginal microbiome from DNA isolated from standard cytology / Pap-smear samples. It is particularly useful when measurements of BV are not available, but cervicovaginal DNA isDocket No. 182219.00297

[0235] available (e.g., from HPV testing). Cervicovaginal samples with molBV scores of 0-3 were classified as mBV-negative, those with >3 - <7 were classified as mBV-intermediate, and those with 7-10 were classified as mB V-positive. mBV states determined by molBV were robust across three large independent cohorts spanning US and continental African women as demonstrated by AUCs of 0.88-0.98. (Usyk, M. et al. (2022) Nat. Commun. 13, 233). Cervicovaginal microbiome community state types (CSTs) were generated using VALENCIA. (France, M. T. et al. (2020) Microbiome 8, 166). To incorporate CST definitions into the mBV analysis, mBV-positive (i.e., molBV score 7-10) participants were dichotomized into mBV-A subtype if they concurrently had CST-IVA and into mBV-B if they had an molBV score of 7-10 and not CST-IVA.

[0236] A sensitivity analysis was performed comparing molBV categorical states (i.e., mBV- positive, mBV-intermediate vs. mBV-negative), the clinical Amsel BV diagnosis from the participant clinical visit, and CST distributions. The Amsel diagnosis and CST distributions concurred with the molBV derived mBV-positive and mBV-intermediate categories. Comparison of the models using Akaike Information Criterion (AIC) and Schwarz’s Bayesian Information Criterion (BIC), which measure goodness of fit, indicated that the mBV models had the best fit of the data.

[0237] NGS reads from this study were deposited in the Qiita repository (Gonzalez, A. et al. (2018) Nat. Methods 15, 796-798) (Study ID: 14884).

[0238] Negative controls (n = 17) and positive controls (n = 17) were used to control for contamination and downstream amplification, sequencing, and bioinformatics. The negative controls had an average (SD) read recovery of 2,803 (1,000) reads compared with the true samples which had an average of 36,384 (13,512) reads, (p value < 0.001). For ITS1 sequencing, the recovery was 1,741 (4,563) reads in the negative controls and 7,774 (20,879) in true samples (p < 0.001). The negative control was water. The positive control was purchased from ZymoBIOMICS Microbial DNA Standard (Zymo Research Corp., Irvine, CA) and after amplification and sequencing showed the expected composition including the bacteria-Pseudomonas aeruginosa, Escherichia coli, Salmonella enterica, Lactobacillus fermentum. Enterococcus faecalis, Staphylococcus aureus. Listeria monocytogenes and Bacillus subtilis, and fungi- Saccharomyces cerevisiae and Cryptococcus neoformans.

[0239] Conditional logistic regression models were fitted using the survival package in R to assess the associations between categorical mBV state (mBV-intermediate and mBV-positive vs.Docket No. 182219.00297

[0240] mBV-negative) and CT at index visit (to) and prospectively (using the CVM assessed at study visit (ti), approximately 6 months prior to detection of CT in cases). The odds ratios (ORs) of incident CT based on mBV states were estimated by comparing subjects who were mBV-positive and

[0241]

[0242] E-interm ediate to

[0243]

[0244] -negative subjects, adjusting for current school attendance, presence of oncogenic HR-HPV types (i.e., HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58 and 59, measured by allele-specific oligonucleotide hybridization as described (Schlecht, N. F. et al. (2021) JAMANetw. Open 4, e2121893)) and sexual risk behavior scores. (Gradissimo, A. et al. (2021). Anti-HPV16 Antibody Titers Prior to an Incident Cervical FIPV16 / 31 Infection. Viruses 13, 1548), All confounding variables for the effect between incident CT and CVM were measured at the time of CVM sampling (i.e., at t-i, to, and t i). The continuous sexual risk behavior composite score is a linear combination of the following variables with categorical states shown in parentheses: lifetime vaginal sex partners (1-4, with 4 representing 4 or more partners), recent (past six months) number of sex partners (0-2, with 2 representing 2 or more partners), history of any pregnancy (0-1), emergency contraception use (0-1), condom usage during recent sex (0-1, with 1 representing never or rarely), ever anal sex (0—1), and lifetime number of anal sex partners (0-2, with 2 representing 2 or more partners); higher values indicate higher sexual risk behavior. In terms of microbial measures, a-diversity (i.e., Chaol and Shannon indices) and b-diversity (Jensen Shannon divergence (JSD) distance) were calculated using the phyloseq package (McMurdie, P. J., and Holmes, S. (2013) PLoS One 8, e61217) in R Statistical significance in a-diversity was determined using the Wilcoxon rank sum test in base R, while PERMANOVA was used to assess significance and obtain R2in b-diversity with the vegan R package. (Oksanen, J. etal. (2010) R package version, 1.17-4). Alluvial plots, representing mBV state transitions between ti and to visits were constructed using the ggplot2 package and the geom_alluvium function. For cross-validation, data was randomly split (50:50) between testing and training sets with effects representing the pooled combined effects of replicable taxa across the 10 testing folds. Biomarker discovery was performed using ANCOM, (Lin, H, and Peddada, S D. (2020) Nat. Commun. 11, 3514) a tool designed specifically for identification of differential bacterial taxa in the context of compositional microbiome data using regression, in the testing set and effect estimates for biomarkers significant after adjustment for FDR < 0.05 were modeled in the testing folds with final mixed-effect pooled effects reported.

[0245] To evaluate cross-validated bacterial taxa from the ti visit, Pearson correlation coefficients were calculated using the R stats package. Determination of a microbial risk scoreDocket No. 182219.00297

[0246] (MRS), (Wang, C. et al. (2022) Microbiome 10,121) an approach analogous to polygenic risk score, combines measures of multiple bacteria into a single value. The diversity MRS (i.e., MRSa) was calculated by performing the corresponding a-diversity calculations (i.e., Chaol and Shannon) using the phyloseq package on the subset of bacterial taxa found to be significantly different between cases (i.e., those that acquired CT) vs. controls using ANCOM followed by cross-validation. To calculate the abundance based MRS (i.e., MRSs) the inventors summed the cross-validated bacterial taxa and applied a weighted sum (weight based on OR of taxon with respect to CT acquisition) as described in the original MRS paper, ( / .). The purpose of the score is to measure the effect of a given bacterial network. For evaluation of subsequent CT reinfection rates after the t+i visit, the time to each CT reinfection and Poisson regression was used within the glm function of the stats package with adjustment using the same variables used in the core analysis (i.e., SRBS, HR-HPV positivity, and school attendance).

[0247] In addition to CT reinfection, a post-hoc exploratory analysis was performed on clinical sequelae of CT infection using self-reported incidence of Pelvic Inflammatory Disease (PID) and miscarriages. Regarding self-reported PID, participants were asked on the study questionnaire at each follow- up visit if “During the past 6 months, have you been told by a doctor or health care provider that you have PID.” This was reported as yes / no / don’t know. Information on prior pregnancies was recorded from clinical interviews conducted during the physical examinations at each study visit and included information on spontaneous abortions. This was recorded as the number of pregnancies that resulted in spontaneous abortion since the last study visit. Mediation analysis

[0248] Although sexual behavior was not found to be statistically associated with incident Chlamydia trachomatis (CT) infection in this cohort, it is considered an important risk factor for not only CT infection, but also development of BV.

[0249] The mediate function was employed from the mediation package in R. This analysis was conducted with 1000 bootstrap simulations to estimate the Average Causal Mediation Effect (ACME), Average Direct Effect (ADE), and the proportion of the total effect that is mediated. For each mediation analysis, binary dummy variables were created for the BV categories (mBV- Negative, mBV-In termediate, and mBV-Positive) and modeled the mediator (mBV) and outcome (incident CT infection) using logistic regression. The mediator models included high- risk HPV status and school attendance as covariates. This approach allowed us to quantify the direct and indirect effects of SRBS on CT infection, mediated through BV. The results areDocket No. 182219.00297

[0250] presented below.

[0251] Value Estimate 95% CI Lower 95% CI Upper p-value

[0252] ACME 0.00021 -0.0016 0.000 0.82

[0253] . ADE -0.00053 -0.025 0.020 0.93

[0254] Prop. Mediated -0.67 -0.91 0.750 0.97

[0255] Total Effect -0.00032 -0.025 0.020 0.94

[0256]

[0257] The direct effect of SRBS on CT infection, not mediated by BV, was negligible and not statistically significant (ADE value). The combined direct and indirect effects of SRBS on CT infection were similarly negligible and not statistically significant (total effect). Finally, the proportion of the total effect that is mediated by BV was not significant. These results might be surprising if taken outside the context of the study cohort characteristics, which included sexually active Adolescent and Young Adults (AYAs), evident in the case-control comparisons of sexual behavior (see Table 1 and Table 4). The characteristics of the study cohort allowed us to directly measure the effect of BV on CT while limiting the impact of sexual behavior. This result is important as it provides additional support that sexual behavior was not a factor driving the association with our main exposure (i.e., the CVM), which is important because of the association of sex with both BV and CT.

[0258] Example 2. Study evaluating the cervicovaginal microbiome (CVM) before, during, and after an incident CT infection and subsequent sequelae

[0259] Participant characteristics

[0260] The study design and cohort characteristics are shown in FIG. 1 and Table 1, respectively. The study sample was selected by matching incident CT cases (w = 187) to controls (n = 373) using 1:2 matching on age, enrollment year, study follow-up time, and prior history of CT infection. Overall, the cohort participants were all sexually active, were an average of 20 years old (range 13 -21 years), were attending school, and had similar sexual risk behavior scores (SRBSs) (Gradissimo, A et al. (2021). Anti-HPV16 Antibody Titers Prior to an Incident Cervical HPV16 / 31 Infection, Viruses 13, 1548). (see Table 4 for a comparison of the individual behavioral variables across cases and controls). molBV scores were analyzed (i.e,, a Nugent-like score of 0-10 using a molBV algorithm (Usyk, M et al. (2022) Nat,Docket No. 182219.00297

[0261] Commun. 13, 233) given the well-known association of BV and CT (Bautista, C. T. et al. (2016) Mil. Med. Res. 13, 4). (Table 1). A molecular assessment of the CVM was used since it provides an objective characterization of BV that is independent of whether a participant has BV-like symptoms (McKinnon, L. R. et al. (2019) AIDS Res. Hum. Retroviruses 35, 219-228), Compared with the controls, the estimated average (median) molBV scores were higher among the cases at the cross- sectional t₀ visit. Interestingly, there was also a significant increase in the molBV scores among future cases in the pre-infection t₋₁ visit (p = 0.037), but not in the post-treatment follow-up visit t₊₁ (p = 0.50), indicating that CVM dysbiosis may be a risk factor for CT acquisition. Features of CSTs also showed a dramatic difference between cases and controls at the incident CT t₀ visit and the pre-infection t₊₁ visit.Docket No. 182219.00297

[0262] Table 1. Characteristics of Participants by Visit

[0263] Pre-infection Visit (t₋₁), n=560 Cross-sectional Visit (to), n=560 Post-infection Visit (t₊₁), n=503 Variables CT Cases CT p- CT Cases CT p-value CT Cases CT p- Controls value Controls Controls value N 187 373 187 373 162 342

[0264] Age Mean ± SD (years) 19.7 ±2.32 19.72 ± 0.91 20.43 ± 20.44 ±2.1 0.86 20.94 ± 21.12 ± 0.42

[0265] 2.12 2.31 2.21 2.08 mBV-categorical

[0266] mBV-Negative#41 (22.8%) 115 (32.2%) - 30 (16.5%) 124 (33.8%) 38 (23.8%) 98 (29.3%) - mB V -Intermediate 67 (37.2%) 123 (34.5%) 0.081 60 (33%) 134 (36.5%) 0.019 64 (40%) 111 (33.1%) 0.12 mBV -Positive 72 (40%) 119 (33.3%) 0.029* 92 (50.5%) 109 (29.7%) <0.0001 58 (36.2%) 126 (37.6%) 0.54 Missing* 7 (3.7%) 16 (4.3%) 5 (2.7%) 6 (1.6%) 2 (1.2%) 7 (2.0%) CSTs

[0267] CST-I⁼ 38 (21.1%) 103 (28.9%) - 16 (8.7%) 106 (28.8%) - 25 (15.6%) 73 (21.7%) - CST-III-A 43 (23.9%) 100 (28%) 0.60 46 (25.1%) 91 (24.7%) 1.30E- 54 (33.8%) 98 (29.2%) 0.13

[0268] 04

[0269] CST-III-B 5 (2.8%) 13 (3.6%) 1.00 9 (4.9%) 24 (6.5%) 0.063 5 (3.1%) 19 (5.7%) 0.79 CST-IV-A 32 (17.8%) 36 (10.1%) 0.005 38 (20.8%) 51 (13.9%) 2.22E- 26 (16.2%) 46 (13.7%) 0.18

[0270] 06

[0271] CST-IV-B 62 (34.4%) 105 (29.4%) 0.067 73 (40.4%) 95 (26.1%) 1.94E- 50 (31.2%) 99 (29.8%) 0.21

[0272] 08

[0273] Missing* 7 (3.7%) 16 (4.3%) 5 (2.7%) 6 (1.6%) 2 (1.2%) 7 (2.0%) Sexual Risk Behavior ScoreA, Mean ± 5.71 ± 1.68 5.71 ± 1.89 0.89 6.14 ± 1.48 5.77 ± 1.87 0.07 6.16 ± 1.47 5.98 ± 1.72 0.36

[0274] SD

[0275] High-Risk HPV

[0276] Positive 4 (2.1%) 12 (3.2%) 8 (4.3%) 6 (1.6%) 11 (6.8%) 9 (2.6%) Negative 183 358 (96.8%) 0.60 179 367 (98.4%) 0.082 151 333 (97.4%) 0.047

[0277] (97.9%) (95.7%) (93.2%)

[0278] Currently Attending School

[0279] Yes 137 277 (74.5%) - 130 239 (64.1%) - 106 201 (58.8%) -

[0280]

[0281] (74.1%) (69.5%) (65.4%)Docket No. 182219.00297

[0282] No 48 (25.9%) 95 (25.5%) 0.92 57 (30.5%) 134 (35.9%) 0.22 56 (34.6%) 141 (41.2%) 0.17 Missing 2 (1.1%) 1 (0.3%) 0 (0.0%) 0 (0.0%) 0 (0.0%) 0 (0.0%) Race / Ethnicity

[0283] 80 133 80 133 71 124

[0284] Afri can-Ameri can / not Hi s panic (42.8%) (35.7%) - (42.8%) (35.7%) (43.8%) (36.3%)

[0285] 32 32 28

[0286] Afri can - Ameri can / Hi spani c (17.1%) 49 (13.1%) 0.789 (17.1%) 49 (13.1%) 0.789 (17.3%) 45 (13.2%) 0.778

[0287] 65 145 65 145 54 135

[0288] No Reported Race / Hispanic (34.8%) (38.9%) 0.183 (34.8%) (38.9%) 0.183 (33.3%) (39.5%) 0.104 Other Race / not Hispanic 2 (1.1%) 19 (5.1%) 0.008 2 (1.1%) 19 (5.1%) 0.008 2 (1.2%) 15 (4.4%) 0.059

[0289] Unknown Race 8 (4.3%) 27 (7.2%) 0.126 8 (4.3%) 27 (7.2%) 0.126 7 (4.3%) 23 (6.7%) 0.216 Abbreviations: mBV, molecular bacterial vaginosis based on molBV scores (see STAR Methods); CSTs, community state types; High-risk HPV types: HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58 and 59.

[0290] Continuous value significance assessed using non-parametric Wilcoxon rank sum test; Categorical variable significance assessed using fisher test.

[0291] ASexual risk behavior score is a composite linear measure that incorporates the total number of lifetime vaginal sexual partners, number of vaginal partners in the last 6 months, lifetime number of pregnancies, condom usage during recent sex, anal sex history and history of emergency contraceptive use (higher values indicating higher risk behavior, see methods for full score description and Supplemental Table 1 for comparison of individual behavior variables). * mBV and CSTs are missing for participants who had 16S sequencing depth <10,000 reads

[0292]

[0293] # mBV-negative and CST-I are used as the reference groups for fisher test

[0294] Table 2. Analysis of Incident Chlamydia trachomatis Infection Associated Factors at the Pre-, Incident, and Post-infection Visit A.

[0295] Pre-infection Visit (t₋₁), Cross-sectional Visit (to), Post-infection Visit (t₊₁), n=503 n=560 n=560

[0296] Covariate aOR* (95% CI) p-value aOR* (95% CI) p-value aOR* (95% CI) p-value mBV (ref: mBV-Negative)

[0297] mBV -Intermediate 1.46 (0.90-2.38) 0.12 2.08 (1.20-3.61) 0.01 1.83 (1.09-3.08) 0.02

[0298] 1.62 (1.01-2.59) 0.04 3.66 (2.18-6.13) <0.001 1.26 (0.73-2.18) 0.41

[0299]

[0300] mBV-PositiveDocket No. 182219.00297

[0301] Sexual Risk Behavior Score* 1.00 (0.89-1.11) 0.95 1.13 (1.00-1.29) 0.05 1.14 (1.00-1.3) 0.04 High-Risk HPV (ref: Negative)

[0302] High-Risk HPV-Positive 0.62 (0.20-1.94) 0.41 2.95 (0.98-8.88) 0.05 3.01 (1.18-7.70) 0.02 Currently Attending School (ref: No)

[0303] Yes 0.94 (0.59-1.51) 0.81 1.78 (1.12-2.84) 0.01 1.37 (0.87-2.16) 0.17 Abbreviations as described in Table 1; ^adjusted odds ratio (aOR) based on multivariable conditional logistic regression;#OR is per unit increase in score.

[0304]

[0305] B.

[0306] Pre-infection Visit (ti) Cross-sectional Visit (to) Post-Infection Visit (t+i) Covariate aOR (95% CI) p-value aOR (95% CI) p-value aOR*(95% CI) p-value CST (ref: CST-I-A)

[0307] CST-III-A 1.1 (0.58-2.08) 0.77 2.83 (1.3-6.16) 0.010 2.02 (1.02-4.00) 0.04 CST-III-B 1.15 (0.32-4.12) 0.83 2.58 (0.83-8.06) 0.10 1.18 (0.34-4.1) 0.79 CST-TV-A 2.46 (1.16-5.19) 0.02 6.50 (2.65-15.92) <0.001 2.36 (1.02-5.47) 0.05 CST-IV-B 1.44 (0.77-2.7) 0.25 4.31 (2.04-9.12) <0.001 1.74 (0.85-3.58) 0.13 Sexual Risk Behavior Score* 0.97 (0.85-1.11) 0.63 1.16 (1.00-1.34) 0.05 1.12 (0.97-1.29) 0.12 High-Risk HPV (ref: Negative)

[0308] High-Risk HPV-Positive 0.66 (0.17-2.53) 0.55 1.6 (0.44-5.91) 0.48 2.96 (1.02-8.64) 0.05 Currently Attending School (ref: No)

[0309] Yes 0.97 (0.54-1.75) 0.92 1.39 (0.81-2.36) 0.23 1.48 (0.91-2.42) 0.12 Abbreviations as described in Table 1; *adjusted odds ratio (aOR) based on multivariable conditional logistic regression; #OR is per unit increase in score.

[0310] Post-infection (tn) visit models the outcome of the CVM measured at t+ i following treatment of CT with antibiotics with respect to the

[0311]

[0312] matched controls using the original risk-set sampling designation at toDocket No. 182219.00297

[0313] Table 3. Stratified Molecular BV States and Risk of CT Incident Infection and Reinfection

[0314] (A) Pre-infection Visit (t-i) CVM and (B) Post-infection Visit (t+i) CVM and CT Incident Infection at to Reinfection at to Covariate aOR* (95% CI), p-value aRR (95% CI)A, p-value n=560 n=160

[0315] mBV (ref: mBV-Negative)

[0316] mB V- Interm edi ate 1.41 (0.85-2.33) 0.18 3.17 (1.18-11.06) 0.038

[0317] mBV-A 2.38 (1.21-4.69) 0.012 3.58 (1.16-13.28) 0.034 mBV-B 1.46 (0.85-2.51) 0.17 1.57 (0.44-6.32) 0.49 Sexual Risk Behavior Score 0.98 (0.87-1.1) 0.72 1.33 (1.04-1.74) 0.030 High-Risk HPV (ref Negative)

[0318] High-Ri sk HPV-Positive 0.51 (0.14-1.88) 0.31 1.6 (0.46-4.37) 0.40 Currently Attending School

[0319] (ref: No)

[0320] Yes 0.79 (0.48-1.3) 0.36 1.39 (0.68-3.06) 0.39 Time to CT#(ref: Time0)

[0321] Timel 0.77 (0.45-1.32) 0.35

[0322] Time2 1.34 (0.82-2.2) 0.25

[0323] Abbreviations as described in Tables 1 and 2.

[0324] * Adjusted odds ratio (aOR) based on multivariable conditional logistic regression with incident CT status at visit to as the outcome.AAdjusted rate ratio (aRR), calculated by multivariable Poisson regression with post-t₊₁ CT infection counts as the outcome.

[0325] " Time variable was treated as a piecewise linear model with two knots at 6.3 and 7.5 months (<6.3 months, 6.3 - 7.5 months, and >7.5

[0326]

[0327] months).Docket No. 182219.00297

[0328] mBV-A = mBV-positive and CST-I V-A positive mBV-B = mBV-positive and CST-IV-A negative

[0329]

[0330] ASexual Risk Behavior score as described in Table 1Docket No. 182219.00297

[0331] Table 4. Sexual Risk Behavior Score (SRBS) at tO, related to SRBS scores shown in Tables 1, 2 and 3.

[0332] Variables Cases Controls p-value Number of participants 187 373

[0333] SRBS, Mean± SD 5.71± 1.68 5.71± 1.89 0.89 Lifetime vaginal sex partners, Mean± SD 3.22± 0.96 3.06± 1.07 0.14 Past 6 months vaginal sex partners, Mean± SD 1.39± 0.53 1.33±0.54 0.22 Lifetime number of pregnancies, Mean± SD 0.11± 0.32 0.16± 0.36 0.21 Condom usage

[0334] Never 30 (17.2%) 84 (22.6%)

[0335] Rarely 28 (15.1%) 71 (19.1%) 0.99 Sometimes 42 (22.6%) 76 (20.4%) 0.21 Most Times 39 (21.0%) 63 (16.9%) 0.11 All Time 39 (21.0%) 58 (15.6%) 0.06 Doesn’t Apply 6 (3.2%) 20 (5.4%) 0.81 Missing (n = 2) 1 (0.5%) 1 (0.3%) 0.48 Ever emergency contraceptive

[0336] No 39 (29.5%) 61 (22.7%) 0.14 Yes 93 (70.5%) 208 (77.3%)

[0337] Missing (n = 159) 55 (29.4%) 104 (27.9%) 0.46 Ever Anal sex

[0338] No 65 (59.6%) 133 (57.6%) 0.81 Yes 44 (40.4%) 98 (42.4%)

[0339] Missing (n = 220) 78 (41.7%) 142 (38.1%) 0.61 Table shows the composite sexual risk behavior score (SRBS, line 2) and the component elements broken down. SRBS is a linear summation of the components shown, except for condom use, which is reversed in rank order as described in Gradissimo et al. (2021). Anti-

[0340]

[0341] HPV16 Antibody Titers Prior to an Incident Cervical HPV16 / 31 Infection. Viruses 13, 1548.

[0342] Table 5. Comparison of Total Cohort and Cohort with Follow-up Data, related to Table 3 model (B) (i.e., characteristics of patients with follow-up data vs. full cohort).

[0343] Group with

[0344] Variable Total Group (to) Follow-up (to) p-value Number of participants 560 502

[0345] Case-Control

[0346] Cases 187 (33.4%) 160 (31.9%) 0.60 Controls 373 (66.6%) 342 (68.1%) -

[0347]

[0348] Age, Mean ± SD (years) 20.44± 2.17 20.39± 2.16 0.77Docket No. 182219.00297

[0349] mBV at to

[0350] mBV-Negative 154 (28.1%) 133 (27.1%) 0.14 mBV -Intermediate 194 (35.3%) 174 (35.4%) - mBV-Positive 201 (36.6%) 184 (37.5%) - Missing (n = 11)

[0351] Sexual Risk Behavior Score,

[0352] 5.89 ± 1.76 5.88 ± 1.74 0.90 Mean ± SD

[0353] High-Risk HPV-Positive

[0354] negative 546 (97.5%) 488 (97.2%) 0.85 positive 14 (2.5%) 14 (2.8%) - Currently attending school

[0355] No 191 (34.1%) 170 (33.9%) 0.95 Yes 369 (65.9%) 332 (66.1%) - See Table 1 for abbreviations.

[0356]

[0357] Table 6 Sensitivity Analysis of Incident Chlamydia trachomatis Infection Associated Factors at the Incident Visit comparing three approaches for analyzing the CVM, related to STAR Methods.

[0358] Model Variable coef se(coef) p-value mBV1

[0359] mBV -Intermediate 0.73 0.27 0.0087 mBV-Positive 1.29 0.26 8.32E-07 SRBSA0.12 0.064 0.049 High-Risk HPV 1.08 0.56 0.054 Currently Attending School 0.57 0.23 0.014 Amsel2

[0360] Amsel-Inconclusive 0.24 0.63 0.70 Amsel-BV 1.09 0.39 0.0052 SRBSA0.14 0.069 0.037 High-Risk HPV 0.60 0.63 0.33 Currently Attending School 0.31 0.25 0.22 CST3

[0361] csT.ni 0.97 0.33 0.0037 CST. IV 1.67 0.34 1.43E-06 SRBSA0.12 0.062 0.039 High-Risk HPV 0.96 0.55 0.083

[0362]

[0363] Currently Attending School 0.40 0.22 0.068Docket No. 182219.00297

[0364] Table shows the results of a sensitivity analysis comparing three approaches for analyzing the cervicovaginal microbiome. All three models use incident CT as the outcome while holding SRBS, HR-HPV status, and school attendance constant and varying the CVM / BV measurements at to (i.e., mBV, Amsel, and CSTs) to test consistency across these approaches. See Table 1 for abbreviations.

[0365] Model selection statistics using Akaike Information Criterion (AIC) and Schwarz’s Bayesian Information Criterion (BIC) are shown below for each model (lower value indicates better model fit).

[0366] 1. mBV AIC and BIC values are 365.16 and 381.18, respectively.

[0367] 2. Amsel's AIC and BIC values are 397.81 and 413.96, respectively,

[0368] 3. CST AIC and BIC values are 375.49 and 391.64, respectively.

[0369]

[0370] Risk factors for development of incident CT infection (pre-infection t-i visit)

[0371] To determine features of the CVM that were risk factors for incident CT infection, the different facets of the microbiome were analyzed. In terms of a diversity, bacterial evenness and richness were significantly elevated in AYA women before acquiring CT compared with controls. Similarly, the overall community composition of bacteria as measured by beta diversity also revealed a significant difference prior to the acquisition of CT. However, neither diversity measure for the fungal communities was associated with incident CT. Identifying BV-like features, e.g., increased microbial diversity (Usyk, M. et al, (2022) Nat. Commun. 13, 233; Coleman, J. S., and Gaydos, C. A. (2018). Molecular diagnosis of bacterial vaginosis: an update. J. Clin. Microbiol.

[0372] 56, e00342-00318) and elevated molBV scores at visit t-j, supported the association between BV and prospective risk of CT (detected at visit to). In order to further evaluate BV and risk of incident CT, multivariable conditional logistic regression and categorical states of mBV were used as shown in the pre-infection visit model (t-1 ) in Table 2. An mBV-positive state (corresponding to a molBV score of 7-10) was significantly associated with development of CT (odds ratio [OR]:::1.62, 95% confidence interval [CI]: 1.01-2.59, p::::0.04), whereas mBV-intermediate (molBV score >3-<7) also had an increased risk for CT, but did not achieve statistical significance (OR = 1.46, 95% CI: 0.90–2.38, p = 0.12), There was no evidence that high-risk human papilloma virus (HR-HPV) infection at tj was an independent risk factor for incident CT (p = 0.40). To examine the extent to which the participants’ sexual behavior was acting through BV leading to subsequent incident CT infection, a causal mediation analysis was performed (see “mediation analysis” in the STAR Methods). Results of the mediation analysis revealed that differences in the SRBSs wereDocket No. 182219.00297

[0373] not associated with incident CT, nor was the potential association with SRBSs mediated by BV. This result is likely due to the homogeneity of sexual behaviors in the study cohort.

[0374] Categorizing the CVM using CSTs and mBV (pre-infection t-i visit)

[0375] A complementary approach to characterize the CVM employs 16S rRNA amplicon sequencing to define bacterial cervicovaginal CSTs, as previously described (France, M. T. et al. (2020) Microbiome 8, 166). Table 1 shows the counts and proportions of CSTs across study visits. To incorporate CSTs into a Nugent-like clinical framework of BV, the distribution of molBV scores with CSTs was compared and it was found that the Lactobacillus-dominated CSTs showed consistently low molBV scores, whereas mBV-positives (i.e., molBV 7-10) were primarily limited to CST-IV-A and CST-IV -B. CST- IV -A is characterized by the species Candidatus Lachnocurva vaginae (previously known as BVAB1), whereas CST-IV-B is associated with Atopobium vaginae. Modeling the CSTs as risk factors for incident CT revealed that surprisingly, only CST-IV-A showed a prospective association with incident CT (OR = 2.46, 95% CI: 1, 16—5.19, p = 0,020), and CST-IV-B showed elevated risk but did not reach statistical significance (OR = 1.44, 95% CI: 0.77-2.70, p = 0.25) (Table 2). To incorporate CST definitions into the categorical mBV analysis, mBV-positive participants at the t-i visit were categorized into mBV- A subtype if they concurrently had CST-IV-A and into mBV-B subtype if they did not have CST-IV-A Table 3 (see model i, pre-infection visit) shows the full model for these mBV subtypes. The analysis revealed that only participants with mBV-A showed a statistically significant elevated risk for acquiring a CT infection (OR = 2.38, 95% CI: 1.21–4.69, p = 0.012). Given that mBV-A showed increased risk for CT, the features of this BV subtype compared with mBV -B was further examined using analysis of composition of microbes (ANCOM). Results confirmed that Candidatus Lachnocurva vaginae was the top species associated with this mBV-A, and levels of this species showed an area under the curve (AUC) for classification of 0.93. Direct analysis of the levels of Candidatus Lachnocurva vaginae showed a 33-fold increase of Candidatus Lachnocurva vaginae in mBV-A when compared with mBV-B.

[0376] Prospective risk of CT acquisition using a polymicrobial risk score (pre-infection t-i visit)

[0377] It was determined whether individual taxa acted as independent risk factors or were part of a polymicrobial community for incident CT. First, differential abundance analysis was performedDocket No. 182219.00297

[0378] with ANCOM and 10-fold cross-validation. FIG. 2 shows the composite results for all bacteria found to be significantly associated with prospective CT acquisition. Overall, 10 taxa were shown to be significant markers for CT acquisition, including Candidates Lachnocurva vaginae, Prevotella, Megasphaera, and Clostridium, among other BV-associated bacteria. (Fredricks, D. N. et al. (2005) N. Engl. J. Med. 353, 1899-1911). Matrix analyses suggested that these taxa belonged to a microbial network as they were highly correlated (pairwise correlations > 0.6, p < 0.001). To further investigate how these taxa were collectively related to CT risk, the microbial risk score (MRS) (which is analogous to a polygenic risk score (Wang, C, et al. (2022) Microbiome 10, 121) was employed. The weighted analysis for CT acquisition per unit increase in MRS, OR = 2,50 (95% CI: 1,35- 5.60, p = 0.0037). The values of the MRS risk estimates were considerably higher than individual bacterial risk estimates, indicating the importance of bacterial communities (see FIG. 2).

[0379] Perturbation of the Cervicovaginal Microbiome (CVM) by Chlamydia trachomatis (CT): Cross-Sectional Associations at the Incident Visit (to)

[0380] The CVM bacterial alpha diversity was higher in CT cases compared to controls, whether measured by species richness (Chaol, p = 0.046) or by metrics incorporating evenness (Shannon, p = 2.50 x 10”7). Fungal alpha diversity was not associated with CT case status. Similarly, bacterial beta diversity analysis revealed significant compositional differences associated with CT case status (R2= 0.046, p = 0.001), whereas fungal beta diversity did not show such differences (R2= 0.001, p 0.89).

[0381] Differential abundance analysis at the species level identified 37 bacterial species consistently differing between CT cases and controls. As expected, CT was the most differentially abundant species, with higher levels in cases than controls (W-stat = 41, false discovery rate [FDR] < 0.05), although it accounted for only 0.054% of total microbial biomass at the cross-sectional visit (to). Across all visits, CT reads exhibited a pronounced spike only at the to visit (p::::2.90 x 10−16, with no significant differences at the pre-infection (t-i) or post-treatment (t-n) visits (p = 0.60 and p = 0.86, respectively).

[0382] Based on 16S rRNA V4 amplicon sequencing, CT reads were detected in 100 of 187 cases (53.5%) compared to 61 of 373 controls (16.4%) at the incident infection visit (to), correspondingDocket No. 182219.00297

[0383] to an odds ratio (OR) of 5.24 (95% CI: 3.44-8.05; p = 2.2 × 10−16). Evaluation of bacterial changes associated with CT acquisition revealed six genera (including CT itself) that were consistently different when comparing t-i to to in cases. Additionally, an increase in the molecular bacterial vaginosis (molBV) score was observed among cases during the transition from pre-infection to incident infection (t-i vs. to) (p = 0.018).

[0384] BV states mBV-positive and mBV- intermediate were significantly associated with CT infection at the time of CT detection (to) (OR = 3.66, 95% CI: 2.18-6.13, p < 0.001, and OR = 2,08,95% CI: 1.20-3.61, p = 0.010, respectively) (see Table 2, cross-sectional (to) model). There was also an elevation in the odds of observing multiple elevated CSTs in the CT case group when using CST-I-A as the reference (Table 2), particularly the BV-associated CSTs dominated by either Candidatus Lachnocurva vaginae (CST-IV-A) or Atopobium vaginae (CST-IV-B) (OR = 6,50, 95% CI: 2.65-15,92, p < 0.001 and OR = 4.31, 95% CI: 2.04-9.12, p < 0.001, respectively). Thus, changes in multiple features of the CVM suggest that CT infection has a broad impact on the microbiome composition. There was also a marginally significant increase in the odds of detecting high- risk HPV types (in the mBV model, Table 2) and in the odds of having a higher SBRS (both models, Table 2). School atendance was only significant in the mBV and not the CST model.

[0385] Post-CT-treatment CVM composition (ti visit)

[0386] The dynamics of the CVM after antibiotic treatment were evaluated for CT (t+i), primarily with azithromycin or doxycycline. There were no identifiable differences between the CVM in CT cases after treatment compared with the controls at this time (ti) for either bacteria or fungal alpha- and beta-diversity analyses or specific species using ANCOM. This suggests a resolution to a control-like state at the group level using these measures. However, multivariable modeling indicated that among the cases, the mBV-intennediate group was elevated compared with the controls (OR = 1.83, 95% CI: 1.09-3.08, p = 0.02; see Table 2 “post-infection visit” (t+1). Similar analysis of CSTs at the post- treatment visit (t+i) indicated that CST-III-A and CST-IV -A remained elevated (OR = 2.02, 95% CI: 1.02-4.00, p = 0.04 and OR = 2.36, 95% CI: 1.02-5.37, p = 0.05, respectively, Table 2). In addition, HR-HPV rates remained significantly elevated among women who had been treated for CT (i.e., CT cases at t+1) compared with controls in both models (see Table 2). ANCOM did not identify any significant differences in bacterial species among the caseDocket No. 182219.00297

[0387] group at ti vs. t+i. The compositional change between the pre- and post-CT CVM of cases was similar to the fluctuation of the CVM over this time (~1 year) in the control group (p = 0.2). Taken together, these data suggest that the case group after treatment reverted to a similar CVM composition that was present prior to CT infection. This reversion is particularly noteworthy given the perturbation of the CVM by CT as measured by the Jensen-Shannon divergence (JSD) difference between ti and to compared with the fluctuation of the control group CVM over this similar 6-month period (p = 0.00067). It should be noted that among controls, comparison of the CVM at 6 months (i.e., t-i vs. to) and 12 months (i.e,, ti vs. t+i) significantly varied (median [IQR] JSD difference 0,071 [0.23] and 0.10 [0,34], respectively, p-value = 0,03). This indicates a continuous effect of time on CVM composition even in the absence of a potent perturbation such as infection with CT. In fact, the transition of the CVM from t-i to to was dependent on the mBV state at the ti visit. Controls showed balanced transitions across mBV states during this 6-month window, while cases tended to favor a transition to an mBV-positive state (i.e., the likelihood of being in the mBV- positive state at tO was dependent on baseline Li CVM: greater in mBV- positive, weaker in mBV- intermediate, and weakest (but still stronger than in controls) in mBV- negative.

[0388] Post-CT reinfection and sequelae

[0389] To examine reinfection rates in participants treated for CT and subsequent incident CT infections in the control group, longitudinal data were used that included clinical testing for CT at each 6-month visit. CT follow-up data beyond the t+i visit was available from 502 / 560 participants (89.6%) who were representative of the original set of participants (Table 5). Thirty-two (20.6%) of 155 participants with CT infection (i.e., case group) had a CT reinfection compared with 10 (4.1%) of 246 control participants who had a subsequent incident CT infection (OR = 6.11, 95% CI: 2.82 -14.42, p < 0.001). Among the cases with follow-up, mBV-A was present in 54 at ti and 57 at t+1, with 18 participants having mBV-A at both t-i and t+i (see Venn diagram in FIG. 3). To determine the risk that mBV-A contributed to CT reinfection, a Poisson regression was used that incorporated the total follow-up time as well as numbers of CT reinfections (Table 3; see model B). Among cases, it was observed that the rate of reinfection was increased among participants with mBV-A (adjusted rate ratio [aRR] of 3.58; 95% CI: 1.16-13.28, p = 0.034) and among those with mB V-intermediate (aRR = 3.17; 95% CI: 1.18-11.1, p = 0.038) (see Table 3; model B).Docket No. 182219.00297

[0390] In addition to CT reinfection, a post hoc exploratory analysis was performed on clinical sequelae of CT infection. (Den Heijer et al. (2019). Chlamydia trachomatis and the risk of pelvic inflammatory disease, ectopic pregnancy, and female infertility: a retrospective cohort study among primary care patients. Clin. Infect. Dis. 69, 1517- 1525). Self-report of PID and miscarriages was used. Among the cases, there were 6 participants who had PID, and 15 participants reported a miscarriage. Participants with an mBV-positive CVM at the post-treatment visit (t+1) were more likely, albeit not significantly, to develop PID and / or have miscarriages, OR = 2.93 (95% CI: 0.88-11.17, p = 0.087) and OR = 4.43 (95% CI: 1.03-30.63, p =0.070), respectively. In terms of CT reinfection and sequelae, significant associations with PID (OR = 1.83, 95% CI: 0.14-13.64, p = 0.61) or miscarriages (OR = 2.33, 95% CI: 0.29-107.93, p = 0.68) were not observed.

[0391] Discussion

[0392] This study prospectively examined the natural history of incident CT infection in a large cohort of primarily Black and Hispanic adolescent and young adult females, who have a 5-fold higher risk of CT acquisition compared with their White counterparts but represent only a small fraction of women studied. (Huai, P. et al. (2020). Prevalence of genital Chlamydia trachomatis infection in the general population: a meta- analysis. BMC Infect. Dis, 20, 589; Torrone, E. et al; Centers for Disease Control and Prevention (CDC) (2014). Prevalence of Chlamydia trachomatis genital infection among persons aged 14-39 years United States, 2007-2012, MMWR Morb, Mortal. Wkly. Rep. 63, 834-838). Cervicovaginal swab samples taken prior to the incident CT infection revealed that participants with mBV had a 62% increased risk for incident CT infection (OR = 1.62, 95% CI: 1.01-2.59, p = 0.04). Further categorization of the mBV-positive state integrating CSTs revealed that a specific subtype of mBV containing CST- IV-A was uniquely associated with a risk of CT acquisition (OR = 2.51, 95% CI: 1.40-4.49, / ? = 0.002). Exploring individual bacterial taxa using 10-fold Monte Carlo cross-validation indicated that a group of 10 highly correlated bacteria (e.g., Candidatus Lachnocurva vaginae, Prevotella, Megasphaera, and Staphylococcus)' collectively increased the risk of CT acquisition. An MRS derived from this set of correlated bacteria indicated the collective risk was greater than the component risk of each bacterium. Cervical samples from the incident visit (to) showed high rates of mBV-positivity with CT and some evidence for co-occurrence of CT with HR-HPVs, although HR-HPV detected atDocket No. 182219.00297

[0393] t-i was not a risk factor for CT. At the time of CT infection, control groups ’ere identified. Analysis of the post-treatment samples indicated that the CVM of CT cases as a group reverted to a compositional state that was nearly identical to the group’s pre-infection CVM. However, there remained a statistically significant elevation of the mBV-intermediate state and elevated levels of CST-IV-A at the post-treatment visit in cases compared with controls. Furthermore, analysis of the post-treatment CVM indicated that the same mBV-positive state (i.e., mBV-A) predicting incident CT infection was also related to the rate of CT reinfection (aRR = 3.58, 95% CI: 1.16-13.28, p = 0.034). These data support the observation that a specific set of bacteria are prospectively associated with risk of CT infection.

[0394] These findings demonstrate that various features of the CVM related to BV (including mBV states, CSTs, and specific taxa) are predictive of CT infection and reinfection, BV is characterized by polymicrobialism (i.e., high bacterial diversity) and a lack of Lactobacillus dominance. This is particularly relevant in the context of CT infection, as Lactobacillus (especially L. crispatus) in the CVM provides protective features by secreting D-lactic acid, which inactivates CT’s infectious particles (elementary bodies) and may mitigate infection. It was further found that specific BV features might increase risk beyond what occurs as a result of simply not having a dominance of Lactobacillus in the CVM. Identifying these granular BV-associated components allows for the identification of individuals at risk and offers means for intervention. This is especially important since some CVM features linked to risk of CT acquisition have also been associated with other health-related conditions. For example, CST-IV-A, which is associated with molecular BV in this study, has been previously linked to persistent BV in a cohort of Kenyan women. If this holds true across populations, certain subtypes of BV may be more likely to persist and simultaneously increase the risk of sexually transmitted conditions such as CT, as shown in this study, or HIV, as observed in Africa. Incorporating BV stratification using specific marker organisms is therefore relevant.

[0395] In this study, BV stratification was introduced by incorporating CSTs into the mBV definition after observing the association of molecular BV and CST-IV-A (but not CST-IV-B) with incident CT infection. It should be noted that not all CST-IV-A cases are linked to BV and confirmed in the studied cohort. In the context of mBV subtypes, mBV-A was characterized by a 33 -fold increase m Candidatus Lachnocurva vaginae compared with mBV-B. This species was theDocket No. 182219.00297

[0396] top taxa associated with CT risk, but its activity was highly correlated with nine other bacteria, suggesting a microbial network effect. Candidatus Lachnocurva vaginae contains the D-lactate dehydrogenase gene (Holm, J. B. et al. (2020) Front. Cell. Infect. Microbiol. 10, 117) which may metabolize D-lactate, potentially weakening D-lactic acid’s protective effect against CT and partially explaining its association with CT risk. While Candidatus Lachnocurva vaginae had the strongest individual association with CT infection, the MRS -derived risk bacteria, including Candidatus Lachnocurva vaginae, Prevotella, and Acinetobacter, showed a significantly stronger combined association with CT acquisition compared with individual taxa. This indicates an additive effect, such as indicating that a microbial network, rather than Candidatus Lachnocurva vaginae alone, is driving the elevated risk, with Candidatus Lachnocurva vaginae possibly serving as a marker rather than a central driver. The post-treatment CVM analysis identified that mBV-A was also associated with CT reinfection, mirroring its association with the incident CT infection. After CT treatment with antibiotics, women exhibited elevated levels of mBV-intermediate and mBV-A states. In fact, prior studies have shown that antibiotic treatment alters CSTs as well. Interestingly, the post-treatment analysis indicated that at the group level, the CVM of cases was similar in overall composition to their baseline CVM However, individual-level analyses revealed that participants shifted into different BV states. Notably, approximately half of the original mBV-A participants had this subtype at both ti and Uj, while others transitioned into mBV-A after antibiotic treatment. This enabled testing whether the mBV-A state was also associated with CT reinfection, which occurs in nearly 25% of incident cases. Chlamydia trachomatis genital tract infections: when host immune response and the microbiome collide. These findings showed that post-treatment mBV-A, mBV-intermediate, and sexual practices measured at the t+i visit were associated with subsequent CT reinfection. Additionally, the CVM profile at the t+1visit indicated an increased risk of PID and miscarriage in cases that had mBV post-CT treatment. These findings indicated that mBV is important in CT infection natural history and its relationship to subsequent sequelae.

[0397] This study presents a molecular analysis of the CVM in the context of incident CT infection in a large Black and Hispanic AYA female population who experience the highest rates of CT infection. By leveraging the routine repeat visits of the Mt. Sinai cohort study and novel molecular assessment of BV, it was demonstrated that AYA women matched in terms of demographics andDocket No. 182219.00297

[0398] behaviors exhibited different risks of CT acquisition based on their CVM. It was further demonstrated that CVM features associated with incident CT infection were also associated with subsequent reinfection. This understanding can form the basis of new public health measures to reduce the burden of CT infections by development of preventive and therapeutic strategies based on more granular features of the CVM

Claims

Docket No. 182219.00297CLAIMSWhat is claimed is:

1. An m vitro method of assessing risk of a disease or disorder in a subject with a vagina and / or cervix, the method comprising:obtaining a vaginal or cervico vaginal sample from the subject;determining, from the sample, a vaginal orcervicovaginal microbiome profile comprising at least a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology; andclassifying the subject as having an elevated risk of disease or disorder when the microbiome profile indicates (i) an mBV-positive state and an elevated level of Candidates Lachnocurva vaginae, or (ii) a microbial risk score (MRS) above a predetermined threshold derived from abundances of a plurality of BV-associated taxa.

2. A method of treating a disease or disorder in a subject with a vagina and / or cervix, the method comprising:obtaining a vaginal or cervicovaginal sample from the subject;determining, from the sample, a cervicovaginal microbiome profile comprising at least a molecular bacterial vaginosis (molBV) score derived from bacterial 16S rRNA gene sequencing data or from an alternative nucleic acid-based technology;classifying the subject as having an elevated risk of disease or disorder when the microbiome profile indicates (i) an mBV-positive state and an elevated level of Candidates Lachnocerva vaginae, or (ii) a microbial risk score (MRS) above a predetermined threshold derived from abundances of a plurality of BV-associated taxa; andwhen the classification indicates elevated risk, administering a therapy according to a dosing regimen effective to reduce a risk of the disease or disorder in the subject.

3. The method of any one of the preceding claims, wherein the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection; Mycoplasma genitaliem infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervicalDocket No. 182219.00297intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

4. The method of claim 2, wherein the therapy comprising a microbiome-targeted therapy that comprises vaginal Lactobacillus probiotic therapy; a non-antibiotic biofilm-disrupting agent; and / or a prebiotic supporting the growth of Lactobacillus and other favorable bacteria,5. The method of any one of the preceding claims, comprising computing the microbial risk score by:performing amplicon sequence variant inference on 16S reads;assigning taxonomy with a trained classifier;transforming taxa abundances via centered log-ratio;computing log-ratios relative to Lactobacillus,' andapplying pre-tramed weights with Platt-scaled calibration to output a probability of incident infection.”6. The method of any one of the preceding claims, wherein taxa abundances used for microbial risk scoring are obtained by targeted quantitative PCR (qPCR), digital PCR, hybrid capture, or other nucleic acid-based quantification methods.

7. The method of any one of the preceding claims, wherein the microbiome profile further comprises a ratio of Candidatus Lachnocurva vaginae to Lactobacillus, and the subject is classified as elevated risk when the ratio exceeds a predetermined cutoff.

8. The method of any one of the preceding claims, wherein the plurality of BV-associated taxa used to compute the microbial risk score comprises at least Candidatus LachnocurvaDocket No. 182219.00297vaginae and one or more of Prevotella, Megasphaera, Clostridium, Staphylococcus, Acinetobacter, or additional taxa selected by differential abundance analysis and cross-validation.

9. The method of any one of the preceding claims, further comprising outputting a report comprising the subject’s risk classification, the mBV category, and the microbial risk score value, and optionally a CST assignment.

10. The method of any one of the preceding claims, wherein the risk assessment is prospective, based on a sample collected prior to detection of incident infection.

11. The method of any one of the preceding claims, wherein the method assesses risk of reinfection following antibiotic treatment, based on a post-treatment sample,12. The method of any one of the preceding claims, wherein the molBV score is computed by transforming 16S rRNA gene amplicon sequence features into a Nugent-like score and classifying into negative, intermediate, or positive categories.

13. The method of claim 8, wherein the CST assignment is performed by a nearest-centroid classification model trained on cervicovaginal community compositions.

14. The method of any one of the preceding claims, wherein the microbial risk score is a weighted sum of taxa abundances or log-ratios relative to Lactobacillus and the weights are derived from odds ratios for incident infection.

15. The method of any one of the preceding claims, wherein the microbial risk score is scaled and categorized using predetermined cutoffs corresponding to increasing odds of incident infection.Docket No. 182219.0029716. The method of any one of the preceding claims, further comprising integrating a sexual risk behavior score and / or high-risk HPV status as covariates into the risk classification model.

17. The method of any one of the preceding claims, wherein the classification step is performed by a trained machine- learning model that outputs a categorical risk label and / or a probability score.

18. The method of any one of the preceding claims, further comprising recommending an intervention selected from follow-up testing, antibiotic stewardship, microbiome-targeted therapy, prebiotic or probiotic therapy, or behavioral counseling based on the elevated risk classification.

19. The method of any one of the preceding claims, wherein the sample is a cervical cytology specimen collected in a liquid-based cytology medium.

20. The method of any one of the preceding claims, wherein the method further comprises detecting high-risk HPV genotypes and reporting the combined microbiome and HPV risk profile.

21. The method of any one of the preceding claims, wherein the subject is classified as elevated risk upon identification of mBV-A.

22. The method of any one of the preceding claims, wherein the elevated risk classification triggers a programmed testing schedule with an increased surveillance interval.

23. The method of any one of the preceding claims, wherein the analysis is performed on archived DNA from a Pap-smear specimen.

24. The method of any one of the preceding claims, wherein the method outputs a longitudinal comparison between a pre-infection sample and a post-treatment sample, highlighting persistence or emergence of the mBV-A subtype.Docket No. 182219.0029725. The method of any one of the preceding claims, wherein the classifier is calibrated such that mBV-intermediate yields an intermediate risk category and mBV-A yields a higher risk category’.

26. The method of any one of the preceding claims, wherein the pipeline flags subjects exhibiting a transition toward mBV-positive states as at risk of incident infection.

27. A method of guiding clinical management for a disease or di sorder, comprising: performing the method of any one of the preceding claims, to identify a subject at elevated risk of the disease or disorder; and recommending at least one of microbiome-targeted therapy, probiotic therapy, or adjusted follow-up intervals.

28. A computer-implemented system for assessing risk of a disease or disorder in a subject with a vagina or cervix, comprising:a data processor and memory storing instructions that, when executed, cause the system to:receive subject-level microbiome data comprising bacterial 16S rRNA gene sequencing features from a vaginal or cervicovaginal sample;compute an molBV score and optionally assign a CST; andcompute a microbial risk score from a set of BV-associated taxa; and classify a risk of a disease or disorder, in the subject based on: (i) mBV-positive with an elevated level of Candidatus Lachnocurva vaginae, or ( ii) the microbial risk score exceeding a threshold; andan interface configured to generate a report comprising the risk classification.

29. The system of claim 28, wherein the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection;Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HI V acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis;Docket No. 182219.00297cervicitis; urethritis with cervical origin; postpartum endometritis; chorioamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

30. The system of claim 28, wherein taxa abundances used for microbial risk scoring are obtained by targeted quantitative PCR (qPCR), digital PCR, hybrid capture, or other nucleic acid-based quantification methods.

31. The system of claim 28, wherein the instructions further cause the system to compute a ratio of Candidatus Lachnocurva vaginae to Lactobacillus and incorporate the ratio into the risk classification.

32. The system of any one of claims 28-31, wherein the microbial risk score is calculated as a weighted sum of abundances or log-ratios of taxa relative to Lactobacillus, with weights derived from effect sizes for incident infection.

33. The system of any one of claims 28 -32, further comprising a model-training module configured to update classification thresholds or weights based on newly ingested labeled data.

34. A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause a computing device to perform the method of any one of claims 1-26 to identify a subject at elevated risk of the disease or disorder.

35. A kit for assessing risk of a disease or disorder in a subject with a vagina and / or cervix, comprising:reagents for extraction of nucleic acids from a vaginal / cervicovaginal sample;a primer set configured to amplify bacterial 16S rRNA gene sequences from the sample; optionally, a primer set configured to amplify fungal or eukaryotic ITS sequences from the sample;Docket No. 182219.00297one or more positive controls and negative controls for sequencing and analysis quality assurance; andinstructions to access software or an analysis pipeline configured to compute an molBV score, optionally assign a CST, compute a microbial risk score from BV-associated taxa, and classify infection risk based on: (i) mBV-positive with an elevated level of Candidatus Lachnocurva vaginae, or (ii) the microbial risk score exceeding a threshold.

36. The kit of claim 35, wherein the disease or disorder is selected from Chlamydia trachomatis infection; Neisseria gonorrhoeae infection; Trichomonas vaginalis infection;Mycoplasma genitalium infection; pelvic inflammatory disease (PID); incident HIV acquisition risk; HSV-2 seroincidence; persistence of high-risk HPV; cervical intraepithelial neoplasia progression; adverse pregnancy outcomes including preterm birth; recurrent bacterial vaginosis; cervicitis; urethritis with cervical origin; postpartum endometritis; chonoamnionitis; premature rupture of membranes (PROM) or preterm PROM (PPROM); infertility or subfertility including in vitro fertilization failure; spontaneous abortion; stillbirth; low birth weight; preeclampsia; vulvovaginal candidiasis recurrence; urinary tract infection susceptibility; group B Streptococcus colonization risk; and autoimmune diseases or disorders.

37. The kit of any one of claims 35-36, further comprising a sample collection device and transport medium suitable for preserving cervicovaginal specimens for microbiome analysis.

38. The kit of any one of claims 35-37, wherein the software implements a nearest-centroid classifier to assign CSTs and a trained risk model that outputs a risk label or probability for incident or recurrent infection.

39. The kit of any one of claims 35-38, wherein the software enables integration of covariates comprising sexual risk behavior metrics and high-risk HPV status into the risk classification.Docket No. 182219.0029740. The kit of any one of claims 35-39, wherein taxa abundances used for microbial risk scoring are obtained by targeted quantitative PCR (qPCR), digital PCR, hybrid capture, or other nucleic acid-based quantification methods.

41. The kit of any one of claims 35-38, wherein the software reports a ratio of Candidatus Lachnocui’va vaginae to Lactobacillus and compares the ratio to a predetermined cutoff associated with elevated risk.

42. The kit of any one of claims 35-40, further comprising written instructions indicating that elevated risk is determined upon identification of an mBV-positive state with an elevated level of Candidatus Lachnocurva vaginae or a microbial risk score above a threshold, and recommending confirmatory testing or clinical follow-up.

43. The kit of any one of claims 35-41, wherein the positive control comprises a defined mock community including Lactobacillus and selected BV-associated taxa, and the negative control comprises nuclease-free water.

44. The kit of any one of claims 35-43, wherein the positive control comprises a defined mock community including Lactobacillus crispatus, Lactobacillus iners, Candidatus Lachnocurva vaginae, and Prevotella bivia.

45. The kit of any one of claims 35-44, wherein the analysis pipeline computes the molBV score by transforming 16S data into a Nugent-like score on a 0-10 scale and classifies the score into negative, intermediate, or positive categories.

46. The kit of any one of claims 35-45, wherein the pipeline produces both prospective risk for incident infection and risk of reinfection following treatment, based on pre- and posttreatment sampling.Docket No. 182219.0029747. The kit of any one of claims 35-46, wherein the microbial risk score is derived from a validated set of BV-associated taxa including at least Candidatus Lachnocurva vaginae and Prevotella.

48. The kit of any one of claims 35-47, wherein the primer set targets the V1-V3, V3-V4, V4, V4-V5, or V5-V7 region of the bacterial 16S rRNA gene.

Images