作者: Shun Guo , Haoran Zhang , Yunmeng Chu , Qingshan Jiang , Yingfei Ma
DOI: 10.1101/2020.09.06.284885
关键词:
摘要: To identify the microbial markers from complex human gut microbiome for delineating disease-related alteration is of great interest. Here, we develop a framework combining neural network (NN) and random forest (RF), resulting in 40 marker species 90 genes identified metagenomic dataset D1 (185 healthy 183 type 2 diabetes (T2D) samples), respectively. Using these markers, NN model obtains higher accuracy classifying T2D-related samples than machine learning-based approaches. The NN-based regression analysis determines fasting blood glucose (FBG) most significant association factor (P<<0.05) microbiome. Twenty-four that vary little across case control are often neglected by statistic-based methods greatly shift different stages T2D development, implying cumulative effect rather individuals likely drives