When we do mapping studies, the idea is to find genes that contribute quantitatively or qualitatively (if probabilistically) to a trait. This is an important objective because, without identifying the genes, we can't understand what they do. However, at this stage of knowledge, much of what we do remains in the black-box category.
In many if not most instances, especially for complex or quantitative traits, variation in genes identifiable by mapping methods accounts for only a fraction, often a small fraction, of the apparently genetic variation in the population. Referring to Figure 5-1, we search for G, but we know that there may be other genes (again, the ovals on the right) with identifiable effects. That is, most of the pheno-typic correlations among relatives remain unexplained after we statistically remove the effects of the already identified genes. As shown in Figure 5-1, this may be due to the effects of shared environmental factors of various kinds; however, even when genes are involved, the residual genetic correlation is treated perforce as a poly-genic aggregate in the figure. When many genes vary but each contributes only a small amount to variation in a trait, there is almost inevitably a large amount of phenogenetic equivalence in the trait, and the concept of genetic causation becomes almost inherently elusive if not actually ephemeral (not to be exactly repeated in the next generation of organisms).
What mapping studies do is to pick out the genes whose alleles happen to be playing a relatively important role in our particular sample. We sometimes say that, although many factors are at work, we are identifying the "rate limiting" ones. In a metabolic pathway, one substrate is converted by a particular enzyme into the substrate for another enzyme and so on, until the original materials are converted into the phenotype being measured. If in our sample, alleles at one of these genes reduce the substrate concentration for subsequent steps, that allele may have a greater effect on the phenotype than alleles at other genes in the pathway. When this is replicated among samples, populations, or species, it may be that the gene in question really is more important in the normal range of variation than the others (or, perhaps, the least buffered by alternative pathways).
Such a rate-limiting step may appear as a QTL in a given sample; however, we may have a misplaced sense of inherency of genetic effect. If the same alleles are missing or alleles at other genes are newly present in subsequent samples, or environments have changed, the original gene may not have the same kind of effect. For these reasons, the generic term "genomic background" is invoked; that is, one strain of a model system responds differently to a particular genetic manipulation than do others (or families differ in the genes responsible for a trait, as detected by mapping studies).
Gene mapping can be done in various ways not described here, but the idea is generally the same—to find genes not already known to affect a trait. But mapping is an imperfect tool for identifying all such genes or characterizing the genetic architecture.
Was this article helpful?