SPERMSEG: analysis of segregation distortion in single-sperm data.
SPERMSEG: analysis of segregation distortion in single-sperm data.
复制标题
SPERMSEG：单精子数据中分离失真的分析。
DOI：
10.1086/302584
复制
发表时间：
1999
影响因子：
9.8
通讯作者：
McPeek,MS
中科院分区：
文献类型：
--
作者：
McPeek,MS
Single-sperm typing has proved to be a valuable tool for detection of segregation distortion at a variety of loci in males (Williams et al. 1993; Leeflang et al. 1996; Takiyama et al. 1997; Girardet et al. 1998; Grewal et al. 1999; Takiyama et al. 1999). Sperm typing can provide the large sample sizes needed to detect even small deviations from 50: 50 segregation that might arise during meiosis or that are due to differential sperm viability during or immediately after spermiogenesis (Leeflang et al. 1996). Furthermore, the problem of ascertainment bias, a serious concern in family-based studies of segregation distortion, is circumvented by sperm typing. However, in order to analyze such sperm data, one must model experimental errors—such as failure of alleles to amplify to a detectable level, deposit of 0 or 11 sperm in a sample, and contamination by exogenous DNA (Cui et al. 1989). Here I describe SPERMSEG, software programmed in C to analyze segregation in single-sperm data. This likelihood-based software is very flexible, allowing for any number of one-and two-marker data sets from one or more donors, with the capabilities to fit virtually any identifiable submodel of interest, to provide confidence intervals for all parameters, and to perform a wide range of hypothesis tests, including simulation-based goodness-of-fit tests. For a likelihood analysis of segregation distortion using single-sperm data, the basic study design involves one or more two-marker data sets from each of several donors. By a two-marker data set, I mean that, for a given donor, two markers, for which the donor is heterozygous and which are linked to the locus of interest, are typed on each of a number of sperm. The reason two markers are typed is that if only one marker were typed on each sperm, it would not be possible to estimate the error parameters in the sperm-typing model. However, with additional assumptions, one-marker data sets can be included in the analysis, in addition to twomarker data sets. Such additional assumptions could include equality, between one-marker and two-marker data sets, of some of the error parameters. Only markers for which the donor is heterozygous can be included in the SPERMSEG analysis. Data from markers for which the donor is homozygous contain no information on segregation distortion (although they may contain a very small amount of information on the error parameters). Thus, it is assumed that each donor’s sperm are typed only for markers for which the donor is heterozygous, with those markers allowed to differ among donors, and with possibly different pairs of markers typed for different subsets of sperm from the same donor. Let G be the locus of interest, with alleles G and g in a given donor. Each two-marker data set involves sperm typed at a pair of markers A and B, at which a given donor has alleles A/a and B/b, respectively, linked to G. Assume that the donor haplotypes are known, say GAB/gab, and assume that the three recombination probabilities,, and, between G and A, G and B, v v v
