R/qtl: high-throughput multiple QTL mapping

Abstract

Motivation: R/qtl is free and powerful software for mapping and exploring quantitative trait loci (QTL). R/qtl provides a fully comprehensive range of methods for a wide range of experimental cross types. We recently added multiple QTL mapping (MQM) to R/qtl. MQM adds higher statistical power to detect and disentangle the effects of multiple linked and unlinked QTL compared with many other methods. MQM for R/qtl adds many new features including improved handling of missing data, analysis of 10,000 s of molecular traits, permutation for determining significance thresholds for QTL and QTL hot spots, and visualizations for cis-trans and QTL interaction effects. MQM for R/qtl is the first free and open source implementation of MQM that is multi-platform, scalable and suitable for automated procedures and large genetical genomics datasets.

Availability: R/qtl is free and open source multi-platform software for the statistical language R, and is made available under the GPLv3 license. R/qtl can be installed from http://www.rqtl.org/. R/qtl queries should be directed at the mailing list, see http://www.rqtl.org/list/.

Contact: kbroman@biostat.wisc.edu.

Fig. 1.

Three examples of MQM plots included in R/qtl. (a) Circular genome interaction plot of the Arabidopsis thaliana glucosinolate pathway (Fu et al., 2007). Logarithm of odds (LOD) scores shown at marker positions are scaled (grey circles), with selected cofactors (red circles) and epistasis between multiple cofactors (green and blue splines). (b) Cis–trans plot of significant QTL (squares) showing cis-acting QTL (diagonal) and a trans-band (vertical, chromosome 5) in Caenorhabditis elegans (Li et al., 2006). (c) Three-way comparison of MQM performance in Arabidopsis thaliana (Fu et al., 2007). LOD score increases when cofactors are added manually to the model. Here, adding more than two cofactors does not improve the model any further (as discussed in the online MQM tutorial).