Institute of Medical Informatics and Statistics, Kiel University, University Hospital Schleswig-Holstein, 24105, Kiel, Germany.
Department of Life Sciences and Chemistry, Jacobs University, 28759, Bremen, Germany.
Hum Genet. 2019 Apr;138(4):375-388. doi: 10.1007/s00439-019-01994-x. Epub 2019 Mar 9.
Metabolic coherence (MC) is a network-based approach to dimensionality reduction that can be used, for example, to interpret the joint expression of genes linked to human metabolism. Computationally, the derivation of 'transcriptomic' MC involves mapping of an individual gene expression profile onto a gene-centric network derived beforehand from a metabolic network (currently Recon2), followed by the determination of the connectivity of a particular, profile-specific subnetwork. The biological significance of MC has been exemplified previously in the context of human inflammatory bowel disease, among others, but the genetic architecture of this quantitative cellular trait is still unclear. Therefore, we performed a genome-wide association study (GWAS) of MC in the 1000 Genomes/ GEUVADIS data (n = 457) and identified a solitary genome-wide significant association with single nucleotide polymorphisms (SNPs) in the intronic region of the cadherin 18 (CDH18) gene on chromosome 5 (lead SNP: rs11744487, p = 1.2 × 10). Cadherin 18 is a transmembrane protein involved in human neural development and cell-to-cell signaling. Notably, genetic variation at the CDH18 locus has been associated with metabolic syndrome-related traits before. Replication of our genome-wide significant GWAS result was successful in another population study from the Netherlands (BIOS, n = 2661; lead SNP), but failed in two additional studies (KORA, Germany, n = 711; GENOA, USA, n = 411). Besides sample size issues, we surmise that these discrepant findings may be attributable to technical differences. While 1000 Genomes/GEUVADIS and BIOS gene expression profiles were generated by RNA sequencing, the KORA and GENOA data were microarray-based. In addition to providing first evidence for a link between regional genetic variation and a metabolism-related characteristic of human transcriptomes, our findings highlight the benefit of adopting a systems biology-oriented approach to molecular data analysis.
代谢相干性 (MC) 是一种基于网络的降维方法，可用于解释与人类代谢相关的基因的联合表达。从计算的角度来看，“转录组学”MC 的推导涉及将个体基因表达谱映射到事先从代谢网络（目前是 Recon2）衍生的基因中心网络上，然后确定特定的、特定于谱的子网络的连通性。MC 的生物学意义之前已经在人类炎症性肠病等方面得到了例证，但这种定量细胞特征的遗传结构仍不清楚。因此，我们在 1000 基因组/GEUVADIS 数据（n=457）中对 MC 进行了全基因组关联研究（GWAS），并在染色体 5 上 cadherin 18 (CDH18) 基因的内含子区域发现了一个与单核苷酸多态性 (SNP) 相关的全基因组显著关联（先导 SNP：rs11744487，p=1.2×10）。Cadherin 18 是一种参与人类神经发育和细胞间信号传导的跨膜蛋白。值得注意的是，CDH18 基因座的遗传变异以前与代谢综合征相关特征有关。我们的全基因组显著 GWAS 结果在荷兰的另一项人群研究（BIOS，n=2661；先导 SNP）中得到了复制，但在另外两项研究（德国的 KORA，n=711；美国的 GENOA，n=411）中没有得到复制。除了样本量问题外，我们推测这些不一致的发现可能归因于技术差异。虽然 1000 基因组/GEUVADIS 和 BIOS 的基因表达谱是通过 RNA 测序生成的，但 KORA 和 GENOA 数据是基于微阵列的。除了提供区域遗传变异与人类转录组代谢相关特征之间联系的初步证据外，我们的研究结果还强调了采用系统生物学导向的分子数据分析方法的好处。
