Additional file 1 of Quantitative profiling of the vaginal microbiota improves resolution of the microbiota-immune axis

Supplementary Material 1: Supplementary methods. Figure S1: Variation in total bacterial load and vaginal soluble immune factors across CST-IV subgroups. Figure S2: BV-associated bacteria drive the association between total bacterial load and immune factors within CST-III. Figure S3: Genital immune milieu cluster tightly with vaginal microbiota composition. Figure S4: Genital immune milieu is closely tied to vaginal microbiota composition in an independent, Uganda-based confirmatory cohort. Table S1: Association between vaginal CST and sociodemographic variables. Table S2: The absolute abundance of BV-associated bacteria, including G. vaginalis and F. vaginae, but not L. iners, were positively associated with sE-cad and IL-1α. Table S3: Nugent scores of women misclassified by the logistic regression model predicting Nugent BV with bacterial load in the SWOP cohort. Table S4: CST subgroups of women misclassified by the logistic regression model predicting molecular BV with bacterial load in the SWOP cohort. Table S5: Comparison of linear regression models predicting soluble immune factors with different vaginal microbiota characterization metrics. Table S6: Comparison of sociodemographic characteristics based on availability of complete immune data in the SWOP cohort. Table S7: Association between PAM immune cluster and sociodemographic variables. Table S8: Sociodemographic factors for Uganda-based confirmatory cohort. N = 61. Table S9: Nugent scores of women misclassified by the logistic regression model predicting Nugent BV with bacterial load in the Uganda-based confirmatory cohort. Table S10: Comparison of linear regression models predicting soluble immune factors with different vaginal microbiota characterization metrics in the confirmatory Uganda-based confirmatory cohort. Table S11: Primer and probe sequences for qPCR assays quantifying total bacterial load.

补充材料1：补充方法。
图S1：CST-IV亚型间总细菌载量与阴道可溶性免疫因子的变化。
图S2：细菌性阴道病（BV）相关细菌驱动CST-III型内总细菌载量与免疫因子的关联。
图S3：生殖器免疫微环境与阴道菌群组成紧密聚类。
图S4：在独立的乌干达验证队列中，生殖器免疫微环境与阴道菌群组成密切相关。
表S1：阴道菌群型（CST）与社会人口学变量的关联。
表S2：细菌性阴道病（BV）相关细菌（包括加德纳菌（G. vaginalis）和阴道纤毛菌（F. vaginae），但不包括惰性乳杆菌（L. iners））的绝对丰度与可溶性E-钙黏蛋白（sE-cad）和白细胞介素-1α（IL-1α）呈正相关。
表S3：SWOP队列中，通过细菌载量预测纽金特细菌性阴道病（BV）的逻辑回归模型误分类女性的纽金特评分（Nugent scores）。
表S4：SWOP队列中，通过细菌载量预测分子型细菌性阴道病（BV）的逻辑回归模型误分类女性的CST亚型。
表S5：使用不同阴道菌群表征指标预测可溶性免疫因子的线性回归模型比较。
表S6：SWOP队列中基于完整免疫数据可获得性的社会人口学特征比较。
表S7：PAM免疫簇与社会人口学变量的关联。
表S8：乌干达验证队列的社会人口学因素。N=61。
表S9：乌干达验证队列中，通过细菌载量预测纽金特细菌性阴道病（BV）的逻辑回归模型误分类女性的纽金特评分（Nugent scores）。
表S10：乌干达验证队列中使用不同阴道菌群表征指标预测可溶性免疫因子的线性回归模型比较。
表S11：用于定量总细菌载量的定量聚合酶链式反应（qPCR）检测的引物和探针序列。