Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Meta-Analysis
. 2013 Jun;9(6):e1003491.
doi: 10.1371/journal.pgen.1003491. Epub 2013 Jun 13.

Effectively identifying eQTLs from multiple tissues by combining mixed model and meta-analytic approaches

Jae Hoon Sul  1 , Buhm Han, Chun Ye, Ted Choi, Eleazar Eskin
Affiliations
Meta-Analysis

Effectively identifying eQTLs from multiple tissues by combining mixed model and meta-analytic approaches

Jae Hoon Sul et al. PLoS Genet. 2013 Jun.

Abstract

Gene expression data, in conjunction with information on genetic variants, have enabled studies to identify expression quantitative trait loci (eQTLs) or polymorphic locations in the genome that are associated with expression levels. Moreover, recent technological developments and cost decreases have further enabled studies to collect expression data in multiple tissues. One advantage of multiple tissue datasets is that studies can combine results from different tissues to identify eQTLs more accurately than examining each tissue separately. The idea of aggregating results of multiple tissues is closely related to the idea of meta-analysis which aggregates results of multiple genome-wide association studies to improve the power to detect associations. In principle, meta-analysis methods can be used to combine results from multiple tissues. However, eQTLs may have effects in only a single tissue, in all tissues, or in a subset of tissues with possibly different effect sizes. This heterogeneity in terms of effects across multiple tissues presents a key challenge to detect eQTLs. In this paper, we develop a framework that leverages two popular meta-analysis methods that address effect size heterogeneity to detect eQTLs across multiple tissues. We show by using simulations and multiple tissue data from mouse that our approach detects many eQTLs undetected by traditional eQTL methods. Additionally, our method provides an interpretation framework that accurately predicts whether an eQTL has an effect in a particular tissue.

PubMed Disclaimer

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Figure 1
Figure 1. A simple example showing how gene expression and SNP in multi-tissue eQTL studies are encoded in the mixed model of Meta-Tissue.
This example has five samples (S1, S2, S3, S4, and S5) in three tissues (T1, T2, and T3). The leftmost table shows which tissues are collected from each sample; formula image means gene expression of formula imageth sample in formula imageth tissue, and formula image means the tissue is not collected. In this example, each tissue has gene expression measured in three samples. formula image is a vector containing expression of samples in all tissues; there are a total of 9 gene expression values. In the formula image matrix, formula image denotes genotype of formula imageth sample. The formula image matrix contains three intercepts (formula image) and three formula image for the three tissues. formula image is the random effect of the mixed model, and formula image. formula image is formula image matrix whose entry at formula imageth row and formula imageth column is 1 if the formula imageth and formula imageth entries of formula image are collected from the same individual, and 0 otherwise.
Figure 2
Figure 2. Power comparison between the tissue-by-tissue approach, Meta-Tissue fixed effects model (FE), and Meta-Tissue random effects model (RE) using simulated data.
X-axis indicates the number of tissues having effects out of four tissues, and Y-axis is the power.
Figure 3
Figure 3. The average number of eQTLs that the tissue-by-tissue approach, Meta-Tissue FE, and Meta-Tissue RE recover from three tissues generated from the liver tissue.
The liver tissue has 108 samples from which we simulate three tissues of 36 samples. X-axis indicates the number of tissues having effects out of three tissues. The original liver tissue has 389 eQTLs.
Figure 4
Figure 4. The number of eQTLs detected by the tissue-by-tissue approach (TBT), Meta-Tissue FE, and Meta-Tissue RE in A) four and B) ten tissues of mouse, and the overlap of eQTLs detected by the three methods in C) four and D) ten tissues.
The datasets consist of the gene expression levels from 50 individuals (four tissues) and 22 individuals (ten tissues). We apply a p-value threshold of formula image for Meta-Tissue and a threshold of formula image/the number of tissues for tissue-by-tissue. The Venn diagrams (C and D) show the number of eQTLs detected by either TBT, FE, or RE, by TBT and either of FE and RE, by FE and RE, and by all three methods.
Figure 5
Figure 5. The number of eSNPs and eProbes detected by the tissue-by-tissue (TBT) approach, Meta-Tissue FE, and Meta-Tissue RE in A) four tissues and B) ten tissues of mouse.
We apply a p-value threshold of formula image for Meta-Tissue and a threshold of formula image/the number of tissues for TBT.
Figure 6
Figure 6. The mice were generated by creating a chimera with heterozygous 129/Sv cells in a C56Bl/6J blastocyst.
The chimera was crossed with a wildtype C56Bl/6J to obtain heterozygous KOs and homozygous WTs. The heterozygous KOs were backcrossed to wildtype C56Bl/6J to obtain animals that are 75% C56Bl/6J. The male and female heterozygous KOs are intercrossed and only the resulting wildtype males are used in this study. The complicated structure of the cross is due to the fact that the knockouts were designed to be used subsequently for other studies.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Brem RB, Yvert G, Clinton R, Kruglyak L (2002) Genetic dissection of transcriptional regulation in budding yeast. Science 296: 752–5. - PubMed
    1. Brem RB, Kruglyak L (2005) The landscape of genetic complexity across 5,700 gene expression traits in yeast. Proc Natl Acad Sci U S A 102: 1572–7. - PMC - PubMed
    1. Keurentjes JJB, Fu J, Terpstra IR, Garcia JM, van den Ackerveken G, et al. (2007) Regulatory network construction in arabidopsis by using genome-wide gene expression quantitative trait loci. Proc Natl Acad Sci U S A 104: 1708–13. - PMC - PubMed
    1. Chesler EJ, Lu L, Shou S, Qu Y, Gu J, et al. (2005) Complex trait analysis of gene expression uncovers polygenic and pleiotropic networks that modulate nervous system function. Nat Genet 37: 233–42. - PubMed
    1. Bystrykh L, Weersing E, Dontje B, Sutton S, Pletcher MT, et al. (2005) Uncovering regulatory pathways that affect hematopoietic stem cell function using ‘genetical genomics’. Nat Genet 37: 225–32. - PubMed

Publication types