Margaritte-Jeannin P, Babron M C, Génin E, Eichenbaum-Voline S, Clerget-Darpoux F
Genetic Epidemiology Research Unit, INSERM U155, Paris, France.
Genet Epidemiol. 1997;14(6):669-74. doi: 10.1002/(SICI)1098-2272(1997)14:6<669::AID-GEPI20>3.0.CO;2-N.
In the first part of our study we tested linkage with chromosome 18 markers in a sample of bipolar I sib pairs. We did not obtain evidence for linkage but showed that we could not exclude the presence of a disease locus (having even a non-negligible effect). The limitation of the sib-pair sample size, and consequently of the conclusions, was a result of our care in assuring that the linkage analysis was free of possible errors in the marker allele frequencies. In the second part, we illustrated the possible impact of such heterogeneity in a single data set when applying the multipoint (APM) method. An Amish pedigree included in the study of Berrettini et al. was analyzed under two sets of marker allele frequencies. One set corresponds to estimates from the entire data set and the second to estimates from the Amish pedigree only. Very different values for the APM statistics were obtained. Although the real frequencies are unknown for this family belonging to an isolated population, this example illustrates that heterogeneity in the populations from which familial data are collected may artificially increase evidence for linkage and hinder interpretation of the analysis.
在我们研究的第一部分，我们在一组双相I型同胞对样本中测试了与18号染色体标记的连锁关系。我们没有获得连锁的证据，但表明我们不能排除疾病位点的存在（甚至具有不可忽略的效应）。同胞对样本量的限制以及由此得出的结论，是由于我们谨慎确保连锁分析不存在标记等位基因频率的可能误差。在第二部分，我们阐述了在应用多点（APM）方法时，这种异质性在单个数据集中可能产生的影响。对包含在贝雷蒂尼等人研究中的一个阿米什家系，在两组标记等位基因频率下进行了分析。一组对应于整个数据集的估计值，另一组仅对应于阿米什家系的估计值。获得了非常不同的APM统计值。尽管对于这个属于隔离群体的家族来说，真实频率是未知的，但这个例子说明，收集家族数据的群体中的异质性可能会人为增加连锁证据，并阻碍分析的解释。
