Integration of Multi-omics Data for Prediction of Metabolic Traits

In biomarker research, the goal is to construct an prediction rule on the basis of a small number of predictors. Formally, this means representing a macro-level response as a function of molecular features (DNA variants, transcript or protein abundancies) with minimal error. The aim is to develop a framework for selection of a composite biomarker: an ensemble of small number of predictors, that is able to predict the macro-level response.

To benchmark the process of construction of the composite biomarker, we use a mouse model. Mouse model has an advantage
over human samples, as many confounding factors are controlled. Here we use measurements of 35 murine strains
from the BXD recombinant inbred strain panel exposed to high-fat and chow diets. As explanatory variable set we use molecular profile of liver, and as response variables, we have selected 7 phenotypic traits related to metabolism.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 668858. This work was supported (in part) by the Swiss State Secretariat for Education, Research and Innovation (SERI) under contract number 15.0324-2. The opinions expressed and arguments employed therein do not necessarily reflect the official views of the Swiss Government.