Linkage and Recombination

As noted above, when genes are not genetically linked, alleles at the loci segregate independently from one another. So, if locus 1 has alleles A and a, and if locus 2, not linked to locus 1, has alleles B and b, then four gametes can be formed (AB, Ab, aB, and ab). Each of these four will occur with equal frequency (a 1:1:1:1 ratio), and all possible offspring combinations are expected with equal frequency.

If locus 1 and locus 2 are genetically linked to one another, however, deviations from this 1:1:1:1 ratio will be observed. If A and B begin on the same chromosome, then AB and ab will be more common than either aB or Ab. By counting the number of each type and determining the extent of this deviation, one can estimate the extent of recombination between the two loci: A large deviation means little recombination. The "recombination fraction," expressed as a percentage, is an indirect measure of the distance between the loci and is the basis for the development of genetic maps.

Genetic maps order polymorphic markers by specifying the amount of recombination between markers, whereas physical maps quantify the distances among markers in terms of the number of base pairs of DNA. Although mapping in humans has a relatively recent history, the idea of a linear arrangement of genes on a chromosome was first proposed in 1911 by Thomas Hunt Morgan, who was studying the fruit fly, Drosophila melanogaster. The possibility of a genetic map was first formally investigated by the American geneticist Alfred H. Sturtevant in the 1930s, who determined the order of five markers on the X chromosome in D. melanogaster and then estimated the relative spacing among them.

For small recombination fractions (usually less than 10 percent to 12 percent), the estimate of the recombination fraction provides a very rough estimate of the physical distance. In general, 1 percent recombination is equivalent to about one million base pairs of DNA and is defined as one centimorgan. Physical measurements of DNA are often described in terms of thousands of kilobases. Crossing over does not occur equally at all locations, so estimates of distance from physical and genetic maps of the identical region may vary dramatically throughout the genome.