eQTL epistasis: detecting epistatic effects and inferring hierarchical relationships of genes in biological pathways

Mingon Kang 1, Hyung-Wook Chun 2, Chunyu Liu 3, and Jean Gao 1
1.Department of Computer Science and Engineering, University of Texas at Arlington, Arlington TX 76013, USA
2.Department of Mathematics, University of Texas at Arlington, Arlington TX 76013, USA
3.Department of Psychiatry, University of Illinois at Chicago, Chicago, IL 66012, USA


Motivation: Epistasis, interactions among multiple genetic variants, has emerged to explain the `missing heritability' that an individual genetic effect does not account for by genome-wide association studies (GWAS), and also to understand the hierarchical relationships between genes in the genetic pathways. Identification of expression quantitative trait loci (eQTL) epistasis captures the structure of the genetic architecture of gene expression, which is an intermediate molecular phenotype between genotype and the higher level phenotypes such as human diseases. The deviation in the multiplicative model from Fisher's geometric model is common in detecting the epistatic effects. However, despite the substantial successes of many studies with the model, it often fails to discover the functional dependence between genes on an epistasis study, which is an important role of the epistasis study to infer hierarchical relationships of genes in the biological pathway.

Results: We justify the imperfectness of Fisher's model in the simulation study and its application to the biological data. Then, we propose a novel generic epistasis model that provides a flexible solution for various biological putative epistatic models in practice. A genetic effect size of an individual locus in the proposed model enables one to efficiently characterize the functional dependence between genes. Moreover, we suggest a statistical strategy for determining a recessive or dominant link among epistatic eQTLs to enable the ability to infer the hierarchical relationships. The proposed model is assessed by simulation experiments of various settings and is applied to human brain data regarding schizophrenia.

Contact: or

Supplementary document

  • Supplemenary.pdf
  • Software
    The codes are implemented by MATLAB (version: R2012a)

  • Simulation Study: (Run "DEMO_SIMUL.m" and "DEMO_FDR.m")