Balancing the tradeoff between quantity and quality of phenotypic data is critical in omics studies. Measurements below the limit of quantification (BLQ) are often tagged in quality control fields, but these flags are currently underutilized in human genetics studies. Extreme phenotype sampling is advantageous for mapping rare variant effects. We hypothesize that genetic drivers, along with environmental and technical factors, contribute to the presence of BLQ flags. Here, we introduce “hypometric genetics” (hMG) analysis and uncover a genetic basis for BLQ flags, indicating an additional source of genetic signal for genetic discovery, especially from phenotypic extremes. Applying our hMG approach to n = 227,469 UK Biobank individuals with metabolomic profiles, we reveal more than 5% heritability for BLQ flags and report biologically relevant associations, for example, at APOC3, APOA5, and PDE3B loci. For common variants, polygenic scores trained only for BLQ flags predict the corresponding quantitative traits with 91% accuracy, validating the genetic basis. For rare coding variant associations, we find an asymmetric 65.4% higher enrichment of metabolite-lowering associations for BLQ flags, highlighting the impact of putative loss-of-function variants with large effects on phenotypic extremes. Joint analysis of binarized BLQ flags and the corresponding quantitative metabolite measurements improves power in Bayesian rare variant aggregation tests, resulting in an average of 181% more prioritized genes. Our approach is broadly applicable to omics profiling. Overall, our results underscore the benefit of integrating quality control flags and quantitative measurements and highlight the advantage of joint analysis of population-based samples and phenotypic extremes in human genetics studies.

中文翻译：

---

低测量遗传学：通过结合质量控制标志提高遗传发现的能力


在组学研究中，平衡表型数据的数量和质量之间的权衡至关重要。低于定量限 （BLQ） 的测量值通常在质量控制字段中被标记，但这些标志目前在人类遗传学研究中未得到充分利用。极端表型采样有利于映射罕见的变异效应。我们假设遗传驱动因素以及环境和技术因素导致 BLQ 标志的存在。在这里，我们介绍了“低遗传学”（hMG） 分析并揭示了 BLQ 标志的遗传基础，表明遗传发现的额外遗传信号来源，尤其是来自表型极端的遗传信号。将我们的 hMG 方法应用于 n = 227,469 名具有代谢组学特征的英国生物样本库个体，我们揭示了 BLQ 标志的遗传性超过 5%，并报告了生物学相关的关联，例如，在 APOC3、APOA5 和 PDE3B 基因座。对于常见变异，仅针对 BLQ 标志训练的多基因分数以 91% 的准确率预测相应的数量性状，验证了遗传基础。对于罕见的编码变异关联，我们发现 BLQ 标志的代谢物降低关联的不对称富集度高出 65.4%，突出了推定的功能丧失变异对表型极端值有很大影响的影响。二值化 BLQ 标志和相应的定量代谢物测量的联合分析提高了贝叶斯罕见变异聚集测试的功效，导致优先级基因平均增加 181%。我们的方法广泛适用于组学分析。 总体而言，我们的结果强调了整合质量控制标志和定量测量的好处，并强调了在人类遗传学研究中对基于群体的样本和极端表型进行联合分析的优势。