Matrix eQTL: Ultra fast eQTL analysis via large matrix operations

By Andrey Shabalin,

... just shortened my computing time from a year to a couple of days
Adaikalavan Ramasamy, Ph.D., King's College London external link

Official tool of the GTEx project external link
Used by seeQTL browser external link
Basis of Multiple Tissue eQTL external link

Please cite R package MatrixEQTL as:
Shabalin, A.A. Matrix eQTL: Ultra fast eQTL analysis via large matrix operations. Bioinformatics 28, no. 10 (2012): 1353-1358. external link

Key features

  • Designed for eQTL analysis of large datasets.
  • Performs testing for all or only local transcript-SNP pair.s
  • Ultra-fast, no loss of precision.
  • Equally fast for models with covariates.
  • Supports
    • Linear additive and ANOVA models. Supports testing for the effect of genotype-covariate interaction.
    • Covariates to account for sex, population structure, surrogate variables, etc.
    • Correlated and heteroskedastic errors.
    • Correction for multiple testing using FDR external link.
    • Separate p-value thresholds and FDR control for local and distant eQTLs (more info).
  • Convenient R package at CRAN Repository external link.
Performance comparison:
Method No covar. 10 covar.
Matrix eQTL, Matlabexternal link 11.8 11.8 minutes
Matrix eQTL, Rev Rexternal link 14.6 14.6 minutes
Matrix eQTL, R+GOTOexternal link 19.4 19.4 minutes
Plink external link 9.4 583.3 days
Merlin external link 19.6 20.0 days
R/qtl external link 1.0 4.7 days
snpMatrix external link 3.2 5.1 days
eMap external link 17.8 N/A days
FastMap external link 10.3 N/A hours

Info: Details of the testing procedure.

Fact: Matrix eQTL results match those by other software.

Comparison conducted analyzing CF dataset with 573,337 SNPs and 22,011 transcripts over 840 samples. Tested on a quad-core PC, using additive linear models with zero and with 10 covariates.

Pdf icon Manuscript in Bioinformatics (2012)

R logo icon Getting started with Matrix eQTL

R logo icon Features of Matrix eQTL

help icon Reference manual

help icon Frequently Asked Questions

Questions, comments, concerns?
Contact me: Andrey A. Shabalin

Support Matrix eQTL

Update History:

  • 2014, February 24 — Update. Version 2.1.0
    • Added support for more than 3 ANOVA categories (see modelANOVA).
    • Added option to reduce memory consumption for large gene data sets.
    • Faster cis-only analysis. Much faster for small cisDist and small slice sizes.
    • File names for the output can now also be connections or NULL.
    • Redesign of QQ-plots.
    • The plot function now follows the R convention and uses ylim instead of ymin parameter.
    • Fixed a bug occurring for genes collinear with covariates.
    • Removed extra space in the output header (for linear models with beta estimates)
    • Sets pvbins are now fixed for QQ-plots. This makes aggregation of results easier.
  • Previous updates

By Andrey Shabalin