CAS Key Laboratory of Computational Biology, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, 320 Yue Yang Road, Shanghai, 200031, China.
CAS Key Laboratory of Computational Biology, Shanghai Institute of Nutrition and Health, University of Chinese Academy of Sciences, Chinese Academy of Sciences, 320 Yue Yang Road, Shanghai, 200031, China.
Am J Hum Genet. 2023 Sep 7;110(9):1509-1521. doi: 10.1016/j.ajhg.2023.08.002. Epub 2023 Aug 23.
Understanding dosage sensitivity or why Mendelian diseases have dominant vs. recessive modes of inheritance is crucial for uncovering the etiology of human disease. Previous knowledge of dosage sensitivity is mainly based on observations of rare loss-of-function mutations or copy number changes, which are underpowered due to ultra rareness of such variants. Thus, the functional underpinnings of dosage constraint remain elusive. In this study, we aim to systematically quantify dosage perturbations from cis-regulatory variants in the general population to yield a tissue-specific dosage constraint map of genes and further explore their underlying functional logic. We reveal an inherent divergence of dosage constraints in genes by functional categories with signaling genes (transcription factors, protein kinases, ion channels, and cellular machinery) being dosage sensitive, while effector genes (transporters, metabolic enzymes, cytokines, and receptors) are generally dosage resilient. Instead of being a metric of functional dispensability, we show that dosage constraint reflects underlying homeostatic constraints arising from negative feedback. Finally, we employ machine learning to integrate DNA and RNA metrics to generate a comprehensive, tissue-specific map of dosage sensitivity (MoDs) for autosomal genes.
理解剂量敏感性或为什么孟德尔疾病具有显性与隐性遗传模式对于揭示人类疾病的病因至关重要。以前对剂量敏感性的了解主要基于对罕见的功能丧失突变或拷贝数变化的观察，由于这些变体极其罕见，因此其功能基础仍然难以捉摸。因此，我们旨在系统地量化来自一般人群中顺式调控变异的剂量扰动，以获得基因的组织特异性剂量约束图谱，并进一步探索其潜在的功能逻辑。我们通过功能类别揭示了基因剂量约束的固有差异，其中信号基因（转录因子、蛋白激酶、离子通道和细胞机制）对剂量敏感，而效应基因（转运蛋白、代谢酶、细胞因子和受体）通常对剂量具有弹性。我们表明，剂量约束不是功能可替代性的衡量标准，而是反映了源自负反馈的内在稳态约束。最后，我们采用机器学习将 DNA 和 RNA 指标整合在一起，为常染色体基因生成全面的、组织特异性的剂量敏感性图谱（MoDs）。
