Most inbred laboratory mouse strains are known to have originated from a mixed but limited founder population in a few laboratories^{1, 2}. However, the effect of this breeding history on patterns of genetic variation among these strains and the implications for their use are not well understood. Here we present an analysis of the fine structure of variation in the mouse genome, using single nucleotide polymorphisms (SNPs). When the recently assembled genome sequence from the C57BL/6J strain³ is aligned with sample sequence from other strains, we observe long segments of either extremely high (40 SNPs per 10 kb) or extremely low (0.5 SNPs per 10 kb) polymorphism rates. In all strain-to-strain comparisons examined, only one-third of the genome falls into long regions (averaging >1 Mb) of a high SNP rate, consistent with estimated divergence rates between Mus musculus domesticus and either M. m. musculus or M. m. castaneus. These data suggest that the genomes of these inbred strains are mosaics with the vast majority of segments derived from domesticus and musculus sources. These observations have important implications for the design and interpretation of positional cloning experiments.

1. Whitehead Institute for Biomedical Research and Whitehead/MIT Center for Genome Research, 9 Cambridge Center, Cambridge, Massachusetts 02139, USA
2. School of Veterinary Science, The University of Queensland, Queensland 4072, Australia
3. The Sanger Centre, The Wellcome Trust Genome Campus, Hinxton, Cambridgeshire CB10 1RQ, UK
4. These authors contributed equally to this work

Correspondence to: Mark J. Daly^1,4 Correspondence and requests for materials should be addressed to M.J.D. (e-mail: Email: mjdaly@genome.wi.mit.edu) or K.L.T. (e-mail: Email: kersli@genome.wi.mit.edu).