当前位置: X-MOL 学术Nature › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A reference map of potential determinants for the human serum metabolome
Nature ( IF 50.5 ) Pub Date : 2020-11-11 , DOI: 10.1038/s41586-020-2896-2
Noam Bar 1, 2 , Tal Korem 1, 2, 3, 4, 5 , Omer Weissbrod 1, 2, 6 , David Zeevi 1, 2, 7 , Daphna Rothschild 1, 2 , Sigal Leviatan 1, 2 , Noa Kosower 1, 2 , Maya Lotan-Pompan 1, 2 , Adina Weinberger 1, 2 , Caroline I Le Roy 8 , Cristina Menni 8 , Alessia Visconti 8 , Mario Falchi 8 , Tim D Spector 8 , , Jerzy Adamski 9, 10, 11 , Paul W Franks 12, 13 , Oluf Pedersen 14 , Eran Segal 1, 2
Affiliation  

The serum metabolome contains a plethora of biomarkers and causative agents of various diseases, some of which are endogenously produced and some that have been taken up from the environment 1 . The origins of specific compounds are known, including metabolites that are highly heritable 2 , 3 , or those that are influenced by the gut microbiome 4 , by lifestyle choices such as smoking 5 , or by diet 6 . However, the key determinants of most metabolites are still poorly understood. Here we measured the levels of 1,251 metabolites in serum samples from a unique and deeply phenotyped healthy human cohort of 491 individuals. We applied machine-learning algorithms to predict metabolite levels in held-out individuals on the basis of host genetics, gut microbiome, clinical parameters, diet, lifestyle and anthropometric measurements, and obtained statistically significant predictions for more than 76% of the profiled metabolites. Diet and microbiome had the strongest predictive power, and each explained hundreds of metabolites—in some cases, explaining more than 50% of the observed variance. We further validated microbiome-related predictions by showing a high replication rate in two geographically independent cohorts 7 , 8 that were not available to us when we trained the algorithms. We used feature attribution analysis 9 to reveal specific dietary and bacterial interactions. We further demonstrate that some of these interactions might be causal, as some metabolites that we predicted to be positively associated with bread were found to increase after a randomized clinical trial of bread intervention. Overall, our results reveal potential determinants of more than 800 metabolites, paving the way towards a mechanistic understanding of alterations in metabolites under different conditions and to designing interventions for manipulating the levels of circulating metabolites. The levels of 1,251 metabolites are measured in 475 phenotyped individuals, and machine-learning algorithms reveal that diet and the microbiome are the determinants with the strongest predictive power for the levels of these metabolites.

中文翻译:

人血清代谢组潜在决定因素的参考图

血清代谢组包含大量的生物标志物和各种疾病的病原体,其中一些是内源性产生的,一些是从环境中吸收的 1 。特定化合物的来源是已知的,包括高度可遗传的代谢物 2、3 或受肠道微生物组 4、生活方式选择(如吸烟 5 或饮食 6)影响的代谢物。然而,大多数代谢物的关键决定因素仍然知之甚少。在这里,我们测量了来自 491 个人的独特且深度表型的健康人类队列的血清样本中 1,251 种代谢物的水平。我们应用机器学习算法,根据宿主遗传学、肠道微生物组、临床参数、饮食、生活方式和人体测量数据来预测被拒个体的代谢物水平,并获得了超过 76% 的分析代谢物的统计显着预测。饮食和微生物组具有最强的预测能力,每一种都可以解释数百种代谢物——在某些情况下,可以解释超过 50% 的观察到的差异。我们通过在两个地理上独立的队列 7 、 8 中显示出高复制率来进一步验证与微生物组相关的预测,这在我们训练算法时是不可用的。我们使用特征归因分析 9 来揭示特定的饮食和细菌相互作用。我们进一步证明了其中一些相互作用可能是因果关系,因为我们预测与面包呈正相关的一些代谢物在面包干预的随机临床试验后被发现增加。总体,我们的研究结果揭示了 800 多种代谢物的潜在决定因素,为从机制上理解不同条件下代谢物的变化以及设计干预循环代谢物水平的干预措施铺平了道路。在 475 个表型个体中测量了 1,251 种代谢物的水平,机器学习算法显示饮食和微生物组是对这些代谢物水平具有最强预测能力的决定因素。
更新日期:2020-11-11
down
wechat
bug