The advent of metagenomics in animal breeding poses the challenge of statistically modelling the relationship between the microbiome, the host genetics and relevant complex traits. A set of structural equation models (SEMs) of a recursive type within a Markov chain Monte Carlo (MCMC) framework was proposed here to jointly analyse the host-metagenome-phenotype relationship. A non-recursive bivariate model was set as benchmark to compare the recursive model. The relative abundance of rumen microbes (RA), methane concentration (CH4 ) and the host genetics was used as a case of study. Data were from 337 Holstein cows from 12 herds in the north and north-west of Spain. Microbial composition from each cow was obtained from whole metagenome sequencing of ruminal content samples using a MinION device from Oxford Nanopore Technologies. Methane concentration was measured with Guardian® NG infrared gas monitor from Edinburgh Sensors during cow's visits to the milking automated system. A quarterly average from the methane eructation peaks for each cow was computed and used as phenotype for CH4 . Heritability of CH4 was estimated at 0.12 ± 0.01 in both the recursive and bivariate models. Likewise, heritability estimates for the relative abundance of the taxa overlapped between models and ranged between 0.08 and 0.48. Genetic correlations between the microbial composition and CH4 ranged from -0.76 to 0.65 in the non-recursive bivariate model and from -0.68 to 0.69 in the recursive model. Regardless of the statistical model used, positive genetic correlations with methane were estimated consistently for the seven genera pertaining to the Ciliophora phylum, as well as for those genera belonging to the Euryarchaeota (Methanobrevibacter sp.), Chytridiomycota (Neocallimastix sp.) and Fibrobacteres (Fibrobacter sp.) phyla. These results suggest that rumen's whole metagenome recursively regulates methane emissions in dairy cows and that both CH4 and the microbiota compositions are partially controlled by the host genotype.

中文翻译：

---

解开微生物群与复杂性状之间生物学关系的结构方程模型：以奶牛产甲烷为例

宏基因组学在动物育种中的出现对微生物组、宿主遗传学和相关复杂性状之间的关系进行统计建模提出了挑战。这里提出了一组在马尔可夫链蒙特卡罗 (MCMC) 框架内递归类型的结构方程模型 (SEM)，以共同分析宿主-宏基因组-表型关系。将非递归双变量模型设置为基准以比较递归模型。瘤胃微生物 (RA) 的相对丰度、甲烷浓度 (CH4) 和宿主遗传学被用作研究案例。数据来自西班牙北部和西北部 12 个牛群的 337 头荷斯坦奶牛。使用牛津纳米孔技术公司的 MinION 设备从瘤胃内容物样品的全宏基因组测序中获得每头奶牛的微生物组成。在奶牛访问挤奶自动化系统期间，使用 Edinburgh Sensors 的 Guardian® NG 红外气体监测器测量甲烷浓度。计算每头奶牛的甲烷喷出峰的季度平均值并将其用作 CH4 的表型。在递归和双变量模型中，CH4 的遗传力估计为 0.12 ± 0.01。同样，对分类群相对丰度的遗传力估计在模型之间重叠，范围在 0.08 和 0.48 之间。微生物组成与 CH4 之间的遗传相关性在非递归双变量模型中介于 -0.76 至 0.65 之间，在递归模型中介于 -0.68 至 0.69 之间。无论使用何种统计模型，对于与纤毛虫门有关的七个属，均一致地估计与甲烷的正遗传相关性，以及属于 Euryarchaeota (Methanobrevibacter sp.)、Cytridiomycota (Neocallimastix sp.) 和 Fibrobacteres (Fibrobacter sp.) 门的那些属。这些结果表明，瘤胃的整个宏基因组递归地调节奶牛的甲烷排放，并且 CH4 和微生物群组成部分受宿主基因型控制。