Introduction

So far in our examination of the inheritance and evolution of quantitative genetics, we've been satisfied with a purely statistical description of how the phenotypes of parents are related to the phenotypes of their offspring. We've made pretty good progress with that. We know how to partition the phenotypic variance into genetic and phenotypic components and how to partition the genetic variance into additive and dominance components. We know how to predict the degree of resemblance among relatives for any particular trait in terms of the genetic components of variance. We know how to predict how a trait will respond to natural selection.

That's not bad, but in the last fifteen or twenty years the emergence of molecular technologies that allow us to identify large numbers of Mendelian markers has led to a new possibility. It is sometimes possible to identify the chromosomal location, at least roughly, of a few genes that have a large effect on the expression of a trait by associating variation in the trait with genotypic differences at loci that happen to be closely linked to those genes. A locus identified in this way is referred to as a quantitative trait locus, and the name given to the approach is QTL mapping.¹

The basic ideas behind QTL mapping are actually very simple, although the implementation of those ideas can be quite complex. In broad outline, this is the approach:

• Produce a set of progeny of known parentage. One common design involves first crossing a single pair of ``inbred'' parents that differ in expression of the quantitative trait of interest and then crossing the $F_1$s, either among themselves to produce $F_2$s (or recombinant inbred lines) or backcrossing them to one or both parents.

• Construct a linkage map for the molecular markers you're using. Ideally, you'll have a large enough number of markers to cover virtually every part of the genome.²

• Measure the phenotype and score the genotype at every marker locus of every individual in your progeny sample.

• Collate the data and analyze it in a computer package like QTL Cartographer to identify the position and effects of QTL associated with variation in the phenotypic trait you're interested in.

If that sounds like a lot of work, you're right. It is. But the results can be quite informative, because they allow you to say more about the genetic influences on the expression of the trait you're studying than a simple parent-offspring regression.