Data for "Quantitative Trait Locus Mapping Methods for Diversity Outbred Mice"


The search for genes underlying complex phenotypes has been greatly aided by genetic mapping in the mouse and other model organisms. Most genetic mapping studies have been conducted on single generation crosses between two inbred strains. This has resulted in poor mapping resolution and has limited opportunities for discovery due to lack of genetic diversity between the parental strains. Multiparent outbreeding populations have been developed to address these shortcomings of traditional cross designs. Multiparent crosses present new analytical challenges that need to be addressed before their full benefit can be realized. Each genetically unique animal in an outbreeding population must be genotyped using a high-density marker set; regression models for mapping must accommodate multiple founder alleles; and complex breeding designs give rise to polygenic covariance among related animals that must be accounted for in mapping analysis. The Diversity Outbred (DO) mice derive from a multiparent cross that combines the genetic diversity of eight founder strains and they have been maintained for more than 15 generations as a large population with randomized mating. We present a complete analytical pipeline for genetic mapping in DO mice. Here we describe algorithms for probabilistic reconstruction of founder haplotypes using either genotyping array intensity data or genotype call data. We propose several regression models for linkage and association mapping. Simulations are used to evaluate these models to validate a permutation testing strategy. Power analysis suggests that loci with large effects can be detected with as few as 100 DO mice, but traits that account for less than 5% of the variance may require as many as 1,000 animals. The methods described here are implemented in the freely available R package DOQTL.


Use of these data should cite the following references:

Quantitative Trait Locus Mapping Methods for Diversity Outbred Mice
Daniel M. Gatti, Karen L. Svenson, Andrey Shabalin, Long-Yang Wu, William Valdar, Petr Simecek, Neal Goodwin, Riyan Cheng, Daniel Pomp, Abraham Palmer, Elissa J. Chesler, Karl W. Broman, Gary A. Churchill
G3 (Bethesda). 2014 Sep 18;4(9):1623-33. doi: 10.1534/g3.114.013748.



Phenotype Data

  • This file contains the white blood cell and neutrophil counts for 742 DO mice. It is a tabular, comma separated file with column names: neutrophils.csv

Reconstructed Genotype Data for Gatti et al. G3 2014

  • Each file is a binary Rdata file that contains a matrix of genotype probabilities. Dimensions are number of markers (rows) by 36 genotypes (columns).


QTL Results for Gatti et al. G3 2014

Power Simulations