Tapper W, Collins A, Gibson J, Maniatis N, Ennis S, Morton N E
Human Genetics Division, University of Southampton, Southampton General Hospital, Southampton SO16 6YD, United Kingdom.
Proc Natl Acad Sci U S A. 2005 Aug 16;102(33):11835-9. doi: 10.1073/pnas.0505262102. Epub 2005 Aug 9.
Two genetic maps with additive distances contribute information about recombination patterns, recombinogenic sequences, and discovery of genes affecting a particular phenotype. Recombination is measured in morgans (w) over a single generation in a linkage map but may cover thousands of generations in a linkage disequilibrium (LD) map measured in LD units (LDU). We used a subset of single nucleotide polymorphisms from the HapMap Project to create a genome-wide map in LDU. Recombination accounts for 96.8% of the LDU variance in chromosome arms and 92.4% in their deciles. However, deeper analysis shows that LDU/w, an estimate of the effective bottleneck time (t), is significantly variable among chromosome arms because (i) the linkage map is approximated from the Haldane function, then adjusted toward the Kosambi function that is more accurate but still exaggerates w for all chromosomes, especially shorter ones; (ii) the non-pseudoautosomal region of the X chromosome is subject to hemizygous selection; and (iii) at resolution less than approximately 40,000 markers per w, there are indeterminacies (holes) in the LD map reflecting intervals of very high recombination. Selection and stochastic variation in small regions must have effects, which remain to be investigated by comparisons among populations. These considerations suggest an optimal strategy to eliminate holes quickly, greatly enhance the resolution of sex-specific linkage maps, and maximize the gain in association mapping by using LD maps.
两个具有加性距离的遗传图谱提供了有关重组模式、重组基因序列以及影响特定表型基因发现的信息。在连锁图谱中,重组以摩根(w)为单位在单代中进行测量,但在以连锁不平衡(LD)单位(LDU)测量的连锁不平衡(LD)图谱中,可能涵盖数千代。我们使用了国际人类基因组单体型图计划(HapMap计划)中的单核苷酸多态性子集来创建全基因组LDU图谱。重组占染色体臂LDU方差的96.8%,在其十分位数中占92.4%。然而,更深入的分析表明,LDU/w(有效瓶颈时间(t)的估计值)在染色体臂之间存在显著差异,原因如下:(i)连锁图谱是根据霍尔丹函数近似得出的,然后朝着更准确但仍对所有染色体(尤其是较短的染色体)夸大w的科桑比函数进行调整;(ii)X染色体的非拟常染色体区域受到半合子选择;(iii)在每w小于约40,000个标记的分辨率下,LD图谱中存在不确定性(空洞),反映了非常高重组率的区间。小区域内的选择和随机变异必定有影响,这有待通过群体间比较进行研究。这些考虑因素提出了一种最佳策略,即通过使用LD图谱快速消除空洞、极大提高性别特异性连锁图谱的分辨率,并最大化关联图谱中的收益。