Application of structural equation models to construct genetic networks using differentially expressed genes and single-nucleotide polymorphisms

Abstract

Understanding the genetic basis of human variation is an important goal of biomedical research. In this study, we used structural equation models (SEMs) to construct genetic networks to model how specific single-nucleotide polymorphisms (SNPs) from two genes known to cause acute myeloid leukemia (AML) by somatic mutation, runt-related transcription factor 1 (RUNX1) and ets variant gene 6 (ETV6), affect expression levels of other genes and how RUNX1 and ETV6 are related to each other. The SEM approach allows us to compare several candidate models from which an explanatory genetic network can be constructed.