Fine Genetic Mapping

An advanced intercrossed line (AIL) is an experimental population that can provide more accurate estimates of quantitative trait loci (QTL) map location than conventional mapping populations. An AIL is produced by randomly and sequentially intercrossing a population that initially originated from a cross between two inbred lines or some variant thereof. This provides increasing probability of recombination between any two loci. Consequently, the genetic length of the entire genome is stretched, providing increased mapping resolution. In this way, for example, with the same population size and QTL effect, a 95% confidence interval of QTL map location of 20 cM in the F2 is reduced fivefold after eight additional random mating generations (F,,,). Simulation results showed that to obtain the anticipated reduction in the confidence interval, breeding population size of the AIL in all generations should comprise an effective number of 2 100 individuals. It is proposed that AILS derived from crosses between known inbred lines may be a useful resource for fine genetic mapping. T HE normal range of phenotypic variation in a wide variety of physiological and morphological traits has a polygenic basis and is quantitative in nature. That is, trait variation is determined by a number of loci, with allele substitution at each locus having a relatively small effect on trait value; trait expression is also affected by macroand microenvironmental factors. Traits of this nature are termed “complex traits” or “quantitative traits” and the individual loci affecting trait expression are usually termed “quantitative trait loci” (QTL). THODAY (1961) was the first to estimate map location of a QTL using a pair of flanking markers. Since then, several statistical methods have been developed that can exploit the information provided by larger numbers of markers and more complete genome maps. In the main, these methods are based on regression analysis and maximum likelihood, and they provide statistical tests for the presence of a QTL and estimates of the QTL parameters ( i .e . , gene effect, dominance and map location), (WELLER 1986; JENSEN 1989; LANDER and BOTSTEIN 1989; KNAPP et nl. 1990; HALEY and KNOTT 1992; DARVASI et al. 1993;JANSEN 1993; ZENC 1993,1994; JANSEN and STAM 1994). However, these methods are not able to efficiently utilize the increasing ability to saturate a given chromosomal region with very closely spaced markers. In particular, with the usual F P , BC, half-sib or full-sib experimental designs and populations of reasonable size, even when using an infinite Cmrfsponrling author Ariel Darvasi, Department of Genetics, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel. E-mail: [email protected] Genetics 141: 1199-1207 (Novernbel-, 1995) number of markers, a QTL of moderate effect can only be assigned to a map location in a rather broad chromosomal region ( DARVASI et al. 1993). This is due primarily to the lack of sufficient recombinational events in small chromosomal regions, even in large F, or BC popula-