Integration of Multi-Layer Omic Data for Prediction of Disease Risk in Humans

Abstract Text:

Accurate prediction of disease risk is needed for implementing personalized medicine. Despite important advances in the assessment of genetic risk, our ability to predict disease risk based on information from the genome (e.g., SNPs) remains very limited. Owing to developments in high-throughput technologies integrated omic profiles are becoming increasingly available. These data holds information that can be extremely useful for the assessment of disease risk and progression. However, omic data is high dimensional and complex, and we lack a coherent framework for the integration of multi-layer omic data into risk assessment models. In this preceding, we discuss extensions of Whole-Genome Regressions that can be used to incorporate integrated omic profiles for the assessment of disease risk. Some of the models described are evaluated using whole-genome expression profiles for prediction of survival after diagnose of breast cancer.

Keywords: prediction of complex traits, diseases risk, omics integration.