THE DETECTION OF LINKAGE IN TETRAD ANALYSIS

Abstract

STABLISHING the existence of linkage between genes at two loci is a E distinct problem from estimating map-distance once genes are known to be linked. This paper considers two questions that concern linkage detection in organisms with genetically analyzable tetrads. First, what criterion is best for establishing linkage when tetrad segregations are analyzedl? And second, what are the advantages of tetrads compared to random single strands for indicating linkage? In classical single-strand analysis, linkage between two genes is *indicated when the ratio of recombination to non-recombination gametes is significantly less than equality. For organisms (chiefly lower plants) where all four products of a single meiosis can be recovered, similar criteria have been adopted for detecting linkage. Ratios of whole tetrads have rarely been used, but segre-gants collected as tetrads have instead been treated as populations of single strands (e.g., POMPER and BURKHOLDER 1949) or of half tetrads (e.g., WHITEHOUSE 1942) for purposes of determining if recombination : non-re-combination numbers differ significantly from the 1 : 1 ratio expected with independent segregation. (By half tetrad is meant specifically a pair of complementary products, A B + ab or A b + aB, originating from the same tetrad. j When segregants that have been obtained as tetrads are treated in this way, two types of error may result. Linkage may either be indicated spuriously where none exists, or linkage may remain concealed that would become apparent if the data were treated more critically. A dihybrid zygote (A a B b) , from gametes A B and ab, can produce any of three possible tetrad types : parental ditype (A B + A B + ab + ab j , non-parental ditype (A b + A b + aB t a B) and tetratype (A B + A b + aB + a b). Observed numbers of these classes are the basic data of tetrad analysis. Parental ditype (PD) and non-parental ditype (N P D) tetrads are equally probable if genes are unlinked, and it will be shown that a significant departure from equality of these two classes is both a reliable and an efficient criterion for establishing linkage when tetrads are used. The advantages of ditype tetrad numbers will be discussed in connection with examples in table 1, where alternative methods are used to examine data from a variety of tetrad segregations. Non-independence of composzent parts of tetrads in testing departures f r o m random. segregation. It is not correct to assume that the two half tetrads from a single meiotic segregation are independent with respect to their recombinant GENETICS 38: 187 March 1953.