Equivalence of single- and multilocus markers: Power to detect linkage with composite markers derived from biallelic loci

The reintroduction of biallelic markers, now in the form of single-nucleoti de polymorphisms (SNPs), has again raised concerns about the practicality o f the use of markers with low heterozygosity for genomic screening for comp lex traits, even if thousands of such markers are available. Like the early blood-group markers (e.g., Rh and MNS), tightly linked biallelic SNPs can be combined into composite markers with heterozygosity similar to that of s hort-tandem-repeat polymorphisms. The assumptions that underlie the equival ence between single-locus multiallelic and composite markers are presented. We used computer simulation to determine the power of the Haseman-Elston t est for linkage with composite markers when not all of these assumptions ho ld. The Genometric Analysis Simulation Program was used to simulate continu ous and discrete traits, one single-locus four-allele marker, and six biall elic markers. We studied composite markers created from pairs, trios, and q uartets of biallelic markers in nuclear families and in independent sib pai rs. The power to detect linkage with a two-point approach for composite mar kers and with a multipoint approach that incorporated all six biallelic mar kers was compared with that for a single-locus, four-allele reference marke r. Although the power to detect linkage with a single biallelic marker was considerably less than that of the reference marker, the power to detect li nkage with two- and three-locus composite markers was quite similar to that of the reference marker. The power to detect linkage with four-locus compo site markers was similar to that of a multipoint approach.