Introduction: The study of lipoprotein metabolism at the population level can provide valuable information for the organisation of lipoprotein related processes in the body. To use this information towards interventional hypotheses generation and testing, we need to be able to identify the mechanistic connections among the large number of observed correlations between the measured components of the system.
Objectives: To use population level metabolomics information to gain insight on their biochemical networks and metabolism.
Methods: Genetic and metabolomics information for 230 metabolic measures, predominately lipoprotein related, from a targeted nuclear magnetic resonance approach, in two samples of an established European cohort, totalling more than 9400 individuals analysed using phenotypic and genetic correlations, as well as Mendelian Randomisation.
Results: More than 20,500 phenotypic correlations were identified in the data, with almost 2000 also showing evidence of strong genetic correlation. Mendelian randomisation, provided evidence for a causal effect between 9496 pairs of metabolic measures, mainly between lipoprotein traits. The results provide insights on the organisation of lipoproteins in three distinct classes, the heterogeneity between HDL particles, and the association, or lack of, between CLA, glycolysis markers, such as glucose and citrate, and glycoproteins with lipids subfractions. Two examples for the use of the approach in systems biology of lipoproteins are presented.
Conclusions: Genetic variation can be used to infer the underlying mechanisms for the associations between lipoproteins for hypothesis generation and confirmation, and, together with biological information, to map complex biological processes.
Keywords: ALSPAC; Causality; Genetic correlation; Mendelian randomization; Metabolomics; Systems epidemiology.
© 2021. The Author(s).