Lin Yu, Xu Zhilu, Yeoh Yun Kit, Tun Hein Min, Huang Wenli, Jiang Wei, Chan Francis Ka Leung, Ng Siew Chien
Microbiota I-Center (MagIC), Hong Kong SAR, China.
Center for Gut Microbiota Research, Faculty of Medicine, the Chinese University of Hong Kong, Hong Kong SAR, China.
iScience. 2023 Mar 22;26(4):106476. doi: 10.1016/j.isci.2023.106476. eCollection 2023 Apr 21.
Obesity is associated with altered gut microbiome composition but data across different populations remain inconsistent. We meta-analyzed publicly available 16S-rRNA sequence datasets from 18 different studies and identified differentially abundant taxa and functional pathways of the obese gut microbiome. Most differentially abundant genera (, , , and ) were depleted in obesity, indicating a deficiency of commensal microbes in the obese gut microbiome. From microbiome functional pathways, elevated lipid biosynthesis and depleted carbohydrate and protein degradation suggested metabolic adaptation to high-fat, low-carbohydrate, and low-protein diets in obese individuals. Machine learning models trained on the 18 studies were modest in predicting obesity with a median AUC of 0.608 using 10-fold cross-validation. The median AUC increased to 0.771 when models were trained in eight studies designed for investigating obesity-microbiome association. By meta-analyzing obesity-associated microbiota signatures, we identified obesity-associated depleted taxa that may be exploited to mitigate obesity and related metabolic diseases.
肥胖与肠道微生物群组成的改变有关，但不同人群的数据仍不一致。我们对来自18项不同研究的公开可用16S-rRNA序列数据集进行了荟萃分析，确定了肥胖肠道微生物群中差异丰富的分类群和功能途径。大多数差异丰富的属（、、和）在肥胖中减少，表明肥胖肠道微生物群中共生微生物的缺乏。从微生物群功能途径来看，脂质生物合成增加以及碳水化合物和蛋白质降解减少表明肥胖个体对高脂肪、低碳水化合物和低蛋白质饮食的代谢适应。在这18项研究上训练的机器学习模型在预测肥胖方面表现一般，使用10折交叉验证时的中位数AUC为0.608。当模型在八项旨在研究肥胖与微生物群关联的研究中进行训练时，中位数AUC增加到0.771。通过对肥胖相关微生物群特征进行荟萃分析，我们确定了可能用于减轻肥胖及相关代谢疾病的肥胖相关减少分类群。
