Quantitative Traits

It is important to note that QTLs can influence quantitative traits in a number of different ways. First, variation at a QTL can impact quantitative trait levels. That is, the average or mean of the observed phenotypes for the trait may be different among different genotypes (for example, some genotypes will produce taller organisms than others). This is important because much of the basic theory underlying statistical methods for studying quantitative traits is based on genotypic means. For this reason, most genetic studies focus on quantitative trait means. However, there are a variety of other ways QTLs can influence quantitative traits. For example, it is possible that the trait means are the same among different genotypes but that the variances (the spread on either side of the mean) are not. In other words, variation in phenotypic values may be greater for some genotypes than for others—some genotypes, for example, may give a wider range of heights than others. This is believed to be due to gene-gene and gene-environment interactions such that the magnitude of the effects of a particular environmental or genetic factor may differ across genotypes.

It is also possible for QTLs to influence the relationship or correlation among quantitative traits. For example, the rate at which two proteins bind might be due to variation in the QTLs that code for those proteins. As a final example, QTLs can also impact the dynamics of a trait. That is, change in a phenotype over time might be due to variation at a QTL, such as when blood pressure varies with the age of the individual. Thus, QTLs can affect quantitative trait levels, variability, co-variability, and dynamics.

In addition, each type of QTL effect may depend on a particular genetic or environmental context. Thus, the influence of a particular QTL on quantitative trait levels, variability, covariability, or dynamics may depend on one or more other QTLs (an effect called epistasis or gene-gene interaction) and/or one or more environmental factors. Although such context-dependent effects may be very common, and may play an important role in genetic architecture, they are typically very difficult to detect and characterize. This is partly due to limits of available statistical methods and the availability of large sample sizes.