Molecular Extent of Linkage Disequilibrium

项目2的重点是连锁不平衡的模式across 17 q21，一个12 Mb长的区域的基因组与几个有趣的特点。首先，它是一个相对基因丰富的区域，在最新的组装（2002年11月）中分配了253个基因，尽管有几个区域在几百kb内没有识别的基因。其次，有相当多的证据表明，它包含至少一个区域，将有资格作为一个块与强大的LD延伸跨越许多kb：BRCA 1基因似乎是一个“块“与LD跨越至少81 kb; LD检测到之间的D17 S1787和D17 S943在CEPH家庭的一个更长的距离为8.6 Mb。第三，基于两个大型独立的作图研究（Marshfield和Decode），整个区域的重组频率显示出很大的性别差异：D17 S1818和D17 S1877之间的重组率估计为4.29和4.94 cM，在男性和女性分别为24.01和17.33 cM。最后，尽管17 q在17 q21的两侧有许多低拷贝片段重复（染色体间和染色体内），但该区域相对不含此类重复，并且序列组装相对明确。我们将用超过1000个标记来检查这个区域，这些标记以大约10 kb的间隔间隔，尤其是在基因丰富的区域中间隔更近，而在基因贫乏的区域中覆盖更稀疏。随着推定区块的边界（如果它们存在的话）变得明显，我们将尝试更密集的覆盖范围，以更好地定义不同种群的边界和这些边界的变化。所有标记将在代表至少40个群体的约2500个个体中进行分型。从经验上讲，我们将能够确定LD的模式在不同人群中的普遍性。这将为“单倍型图谱”概念的普遍性提供一个测试。LD的模式在不同的 群体、杂合性水平和群体关系都将被用于设计用于估计导致现代人类LD的多个因素的分析。

Project 2 focuses on the patterns of linkage disequilibrium across17q21, a 12 Mb long region of the genome with several interesting characteristics. First, it is a relatively gene rich region with 253 genes assigned in the latest assembly (Nov, '02), though there are a few regions that have no identified genes across several hundred kb. Second, there is considerable evidence that it contains at least one region that would qualify as a block with strong LD extending across many kb: the BRCA1 gene appears to be a "block"with LD across at least 81 kb; LD was detected across an even longer distance of 8.6 Mb between D17S1787 ad D17S943 in the CEPH families. Third, the recombination frequencies across the region, based on two large independent mapping studies (Marshfield and Decode) show a great sex difference: the recombination rate between D17S1818 and D17S1877, which just flank 17q21, is estimated at 4.29 and 4.94 cM, respectively, in males and at 24.01 and 17.33 cM, respectively, in females. Finally, although 17q has many low copy segmental duplications (inter- and intra-chromosomal) on either side of 17q21, this region is relatively free of such repeats and the sequence assembly is relatively unambiguous. We will examine this region with over 1000 markers spaced at approximately 10 kb intervals with closer spacing in especially gene-rich regions and somewhat sparser coverage in the gene poor regions. As the boundaries (if they exist as such) of putative blocks become evident, we will attempt to have denser coverage to better define the boundaries and the variation in those boundaries in different populations. All markers will be typed in about 2500 individuals representing at least 40 populations. Empirically, we will be able to determine how generalizable the patterns of LD are among the various populations. This will provide one test of the generalizability of the "haplotype map" concept. The patterns of LD in the different populations, the levels of heterozygosity, and the population relationships will all be used in analyses designed to estimate the multiple factors that are the causes of LD in modern humans.