Mapping quantitative trait loci (QTL) in sheep. II. Meta-assembly and identification of novel QTL for milk production traits in sheep

Sheep represent an economically important agricultural resource in the global meat, fibre, and milk production systems of both the developed and developing world. The multi-purpose nature of many sheep breeds and the highly specialised single purpose breeds, demonstrate the versatility and suitability of sheep production in a diverse set of production systems [1]. Sheep milk production represents a specialised commodity which has been developed across many breeding systems in either dual purpose, synthetic composite lines, or specialised dairy breeds such as the Awassi, Chios, Comisana, Lacaune, Laxta and Sarda Breeds [2]. Genetic variation has been reported for most of the major milk traits, and has been successfully exploited in genetic improvement programmes for sheep dairy production [2-6].Over the past few decades, numerous quantitative trait loci (QTL) studies have been conducted on many breeds of livestock to enrich our knowledge on the underlying biology and genetic architecture of complex traits. A general review of QTL mapping can be found in Weller [7]. However for milk production and udder health, fewer QTL studies have been conducted [3,8-12] in dairy sheep in comparison with cattle, as reviewed by Khatkar et al. [13] and summarised on the animal genome website [14]. Analysis of QTL information from diverse sources within a species provides important information on the consistency and utility of QTL across breeds, and is amenable to meta-analysis to reach a consensus on QTL location and effect [13]. Furthermore, cross-species comparative analyses with a species for which many QTL for the same traits have been reported such as milk traits in cattle may provide additional insight to QTL in a closely related species. This may give insight into ancestral genes for economically important traits and adds information to accelerate fine-mapping in the QTL information-sparse species.QTL mapping requires robust phenotypes. In the case of lactation measurements, t