Division of Preventive Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA 02215, USA.
BMC Genet. 2005 Dec 30;6 Suppl 1(Suppl 1):S25. doi: 10.1186/1471-2156-6-S1-S25.
Several simulation studies have suggested that a high-density single-nucleotide polymorphisms (SNPs) marker set may be as useful as a traditional microsatellites (MS) marker set in performing whole-genome linkage analysis. However, very few studies have directly tested the SNPs-based genome-wide scan. In the present study, we compared the linkage results from the SNPs-based scan with a map density of 3-cM spacing with those from the MS scan using a 10-cM marker set among 300 nuclear families each from the Aipotu (AI), Danacaa (DA), and Karangar (KA) populations from the simulated Genetic Analysis Workshop 14 Problem 2 data. We found that information contents obtained from the SNPs scan were somewhat lower than those from the MS scan. However, the linkage results obtained from the two scans showed a high degree of similarity. Both scans identified a similar number of chromosomal regions attaining nominal significance (p < 0.05). Specifically, both scans detected confirmed evidence for linkage (NPL >or= 4.07, p = 2 x 10(-5)) to chromosome 1 in the AI families, chromosomes 1 and 3 in the DA families, and chromosomes 3, 5, and 9 in the KA families. An additional confirmed linkage to chromosome 5 in the AI families was detected only by the MS scan. We also observed slightly wider 1-LOD intervals for more of the SNP peaks than for the MS peaks, which is likely due to lower information contents for the SNPs. Subsequent fine-mapping association analysis further identified 2 to 3 markers significantly associated with disease status in each population; B03T3056, B03T3058, and B05T4139 in the AI population, B03T3056 and B03T3058 in the KA population, and B03T3056, B03T3057, and B03T3058 in the DA population. Among the four markers, three were chosen based on results obtained from the two scans, but one was solely from the SNP scan. In summary, our finding suggests that the SNP-based genome scan has the potential to be as powerful as the traditional MS-based scan and offers good identification of peak location for further fine-mapped association analysis.
几项模拟研究表明，高密度单核苷酸多态性 (SNP) 标记集在进行全基因组连锁分析时可能与传统微卫星 (MS) 标记集一样有用。然而，很少有研究直接测试基于 SNP 的全基因组扫描。在本研究中，我们比较了基于 SNP 的扫描与使用 AI、DA 和 KA 三个群体的 300 个核家族中的每一个的 10-cM 标记集的 MS 扫描的连锁结果。我们发现，来自 SNP 扫描的信息含量略低于来自 MS 扫描的信息含量。然而，两种扫描的连锁结果具有高度的相似性。两种扫描都确定了数量相似的染色体区域达到名义显著水平 (p < 0.05)。具体来说，两种扫描都检测到了 AI 家族中染色体 1、DA 家族中染色体 1 和 3 以及 KA 家族中染色体 3、5 和 9 的连锁证据（NPL > 4.07，p = 2 x 10(-5)）。AI 家族中的 MS 扫描还检测到了对染色体 5 的另一个确认的连锁。我们还观察到，SNP 峰的 1-LOD 间隔比 MS 峰略宽，这可能是由于 SNP 的信息量较低。随后的精细映射关联分析进一步确定了每个群体中与疾病状态显著相关的 2 到 3 个标记；AI 群体中的 B03T3056、B03T3058 和 B05T4139，KA 群体中的 B03T3056 和 B03T3058，以及 DA 群体中的 B03T3056、B03T3057 和 B03T3058。在这四个标记中，有三个是基于两种扫描的结果选择的，但有一个是仅从 SNP 扫描中选择的。总之，我们的发现表明，基于 SNP 的基因组扫描具有与传统的 MS 扫描一样强大的潜力，并为进一步的精细映射关联分析提供了良好的峰位识别。
