eQTL networks unveil enriched mRNA master integrators downstream of complex disease-associated SNPs

Haiquan Li, Nima Pouladi, Ikbel Achour, Vincent Gardeux, Jianrong Li, Qike Li, Hao Helen Zhang, Fernando D. Martinez, Joe G.N. Skip Garcia, Yves A. Lussier

Research output: Contribution to journalArticlepeer-review

8 Scopus citations


The causal and interplay mechanisms of Single Nucleotide Polymorphisms (SNPs) associated with complex diseases (complex disease SNPs) investigated in genome-wide association studies (GWAS) at the transcriptional level (mRNA) are poorly understood despite recent advancements such as discoveries reported in the Encyclopedia of DNA Elements (ENCODE) and Genotype-Tissue Expression (GTex). Protein interaction network analyses have successfully improved our understanding of both single gene diseases (Mendelian diseases) and complex diseases. Whether the mRNAs downstream of complex disease genes are central or peripheral in the genetic information flow relating DNA to mRNA remains unclear and may be disease-specific. Using expression Quantitative Trait Loci (eQTL) that provide DNA to mRNA associations and network centrality metrics, we hypothesize that we can unveil the systems properties of information flow between SNPs and the transcriptomes of complex diseases. We compare different conditions such as naïve SNP assignments and stringent linkage disequilibrium (LD) free assignments for transcripts to remove confounders from LD. Additionally, we compare the results from eQTL networks between lymphoblastoid cell lines and liver tissue. Empirical permutation resampling (p<0.001) and theoretic Mann-Whitney U test (p<10-30) statistics indicate that mRNAs corresponding to complex disease SNPs via eQTL associations are likely to be regulated by a larger number of SNPs than expected. We name this novel property mRNA hubness in eQTL networks, and further term mRNAs with high hubness as master integrators. mRNA master integrators receive and coordinate the perturbation signals from large numbers of polymorphisms and respond to the personal genetic architecture integratively. This genetic signal integration contrasts with the mechanism underlying some Mendelian diseases, where a genetic polymorphism affecting a single protein hub produces a divergent signal that affects a large number of downstream proteins. Indeed, we verify that this property is independent of the hubness in protein networks for which these mRNAs are transcribed. Our findings provide novel insights into the pleiotropy of mRNAs targeted by complex disease polymorphisms and the architecture of the information flow between the genetic polymorphisms and transcriptomes of complex diseases.

Original languageEnglish (US)
Pages (from-to)226-234
Number of pages9
JournalJournal of Biomedical Informatics
StatePublished - Dec 1 2015


  • Big data
  • Centrality
  • Complex disease
  • Complex diseases
  • Computational biology
  • Computational genomics
  • Computational medicine
  • Genetics
  • Genomics
  • MRNA
  • Master integrator
  • Network biology
  • Signal integration
  • Single Nucleotide Polymorphism (SNP)
  • Systems biology
  • Systems medicine
  • Transcriptome
  • Translational bioinformatics
  • eQTL

ASJC Scopus subject areas

  • Computer Science Applications
  • Health Informatics


Dive into the research topics of 'eQTL networks unveil enriched mRNA master integrators downstream of complex disease-associated SNPs'. Together they form a unique fingerprint.

Cite this