Vigorito Elena, Lin Wei-Yu, Starr Colin, Kirk Paul D W, White Simon R, Wallace Chris
MRC Biostatistics Unit, University of Cambridge, Cambridge, UK.
Cambridge Institute of Therapeutic Immunology and Infectious Disease (CITIID), Jeffrey Cheah Biomedical Centre, Cambridge Biomedical Campus, University of Cambridge, Cambridge, UK.
Nat Comput Sci. 2021 Jun;1:421-432. doi: 10.1038/s43588-021-00087-y. Epub 2021 Jun 24.
Detecting genetic variants associated with traits (quantitative trait loci, QTL) requires genotyped study individuals. Here we describe BaseQTL, a Bayesian method that exploits allele-specific expression to map molecular QTL from sequencing reads (eQTL for gene expression) even when no genotypes are available. When used with genotypes to map eQTL, BaseQTL has lower error rates and increased power compared with existing QTL mapping methods. Running without genotypes limits how many tests can be performed, but due to the proximity of QTL variants to gene bodies, the 2.8% of variants within a 100 kB window that could be tested contained 26% of eQTL detectable with genotypes. eQTL effect estimates were invariably consistent between analyses performed with and without genotypes. Often, sequencing data may be generated in the absence of genotypes on patients and controls in differential expression studies, and we identified an apparent psoriasis-specific eQTL for in one such dataset, providing new insights into disease-dependent gene regulation.
检测与性状相关的基因变异（数量性状基因座，QTL）需要对研究个体进行基因分型。在此，我们描述了BaseQTL，这是一种贝叶斯方法，即使在没有基因型信息的情况下，也能利用等位基因特异性表达从测序读数中定位分子QTL（用于基因表达的eQTL）。与现有的QTL定位方法相比，当与基因型一起用于定位eQTL时，BaseQTL具有更低的错误率和更高的效能。在没有基因型信息的情况下运行会限制可进行的测试数量，但由于QTL变异与基因体的接近性，在100 kB窗口内可测试的2.8%的变异包含了用基因型可检测到的26%的eQTL。在有和没有基因型信息的情况下进行的分析中，eQTL效应估计始终是一致的。通常，在差异表达研究中，患者和对照没有基因型信息时也可能会生成测序数据，我们在一个这样的数据集中鉴定出了一个明显的银屑病特异性eQTL，为疾病依赖性基因调控提供了新的见解。
