Additional file 1 of Determine independent gut microbiota-diseases association by eliminating the effects of human lifestyle factors

Additional file 1: Fig. S1. ROC curves for a selection of the best classification models of eight diseases. Fig. S2. Comparing the differences of AUCs of nine diseases using five feature types after removing probiotics, vitamin B, and vitamin D. Fig. S3. Comparing the differences of AUCs of nine diseases using five feature types after adding gut microbial diversity. Table S1. Frequency of diet and lifestyle factors. Table S2. Basic information of the dataset. Table S3. Comparing AUC values of nine diseases using five feature types. Table S4. AUCs, sensitivity, and specificity for five types of features of the best model selected according to the AUC score. Table S5. Comparing AUC values of nine diseases using five feature types after removing probiotics, vitamin B, and vitamin D. Table S6. Top 10 features after removing probiotics, vitamin B, and vitamin D for all diseases. Table S7. Spearman’s correlations of the human variables and disease. Table S8. Comparing AUC values of nine diseases using five feature types with adding diversity.

附加文件1：图S1。八种疾病最优分类模型的受试者工作特征（Receiver Operating Characteristic，ROC）曲线。图S2。去除益生菌、维生素B及维生素D后，基于五种特征类型比较九种疾病的曲线下面积（Area Under the Curve，AUC）差异。图S3。添加肠道微生物多样性后，基于五种特征类型比较九种疾病的AUC差异。表S1。饮食与生活方式因素的频次分布。表S2。数据集基本信息。表S3。基于五种特征类型比较九种疾病的AUC值。表S4。依据AUC得分筛选得到的最优模型，其五种特征类型对应的AUC、灵敏度（Sensitivity）及特异度（Specificity）。表S5。去除益生菌、维生素B及维生素D后，基于五种特征类型比较九种疾病的AUC值。表S6。去除益生菌、维生素B及维生素D后，所有疾病对应的前10位特征。表S7。人体变量与疾病的斯皮尔曼（Spearman）相关性。表S8。添加多样性后，基于五种特征类型比较九种疾病的AUC值。