Rao D C, Gu C
Department of Psychiatry, Washington University School of Medicine, St. Louis, Missouri 63110, USA.
Adv Genet. 2001;42:487-98. doi: 10.1016/s0065-2660(01)42038-4.
It is emphasized that two types of errors are made in the testing of a hypothesis, false positive (type I) and false negative (type II). Genome-wide scans involving many markers give rise to the problem of multiple testing, which results in an increased number of false positives, thus necessitating a correction in the nominal significance level. While the literature has concentrated reasonably heavily on controlling false positives in genomic scans, the need to control false negatives has been largely neglected. This chapter highlights this need and attempts to strike a balance between the two error types. The need to develop alternative methods for discriminating between false positives and true positives is also stressed.
需要强调的是，在假设检验中会出现两种错误，即假阳性（I型）和假阴性（II型）。涉及许多标记的全基因组扫描会引发多重检验问题，这会导致假阳性数量增加，因此有必要对名义显著性水平进行校正。虽然文献相当多地集中于控制基因组扫描中的假阳性，但控制假阴性的需求在很大程度上被忽视了。本章强调了这一需求，并试图在这两种错误类型之间取得平衡。还强调了开发区分假阳性和真阳性的替代方法的必要性。
