The International Complex Trait Consortium

R/qtl: An extensible QTL mapping environment

Karl W. Broman1, Zaunak Sen2, Hao Wu2, Gary A. Churchill2

1Department of Biostatistics, Johns Hopkins University, Baltimore, MD 21205 USA
2The Jackson Laboratory, Bar Harbor, ME 04609 USA

ABSTRACT
 
Available computer programs for mapping QTLs in experimental crosses are extremely useful for the application of simple (especially single-QTL) models, but it can be difficult or impossible to modify these programs for use with more complex models. We are developing a QTL mapping environment, R/qtl, as an add-on package for the freely available and widely used statistical language/software R (www.R-project.org). The development of this software as an add-on to R allows us to take advantage of the basic mathematical and statistical functions, and powerful graphics capabilities, that are provided with R. Our goal is to make complex QTL mapping methods widely accessible and allow users to focus on modeling rather than computing.

A key component of computational methods for QTL mapping is the hidden Markov model (HMM) technology for dealing with missing genotype data. We have implemented the main HMM algorithms, with allowance for the presence of genotyping errors, for backcrosses, intercrosses, and phase-known four-way crosses. The results of these algorithms are accessible to the user, and the algorithms can be extended for other types of crosses.

The current version of R/qtl includes facilities for estimating genetic maps, identifying genotyping errors, and performing single-QTL and pairwise scans by interval mapping, Haley-Knott regression and imputation, with possible consideration of additional covariates.

Papers

Review of statistical methods for QTL mapping in experimental crosses.
Crossover interface in the mouse.

References

http://biosun01.biostat.jhsph.edu/~kbroman/qtl