Illuminating the distribution of extreme evolutionary constraint in the human genome from fetal demise to severe developmental disorders

摘要 自然选择从影响生存或生存的基因组区域的人群中清除有害突变 生殖能力。最近对大量人群的遗传学研究揭示了持续不断的 这种分布在人类基因组上的有害突变受到限制。大型联合体 研究发现，受严格限制的基因组区域的突变是许多儿童的主要风险因素 发育障碍(DDS)，这表明高度受限的序列可能在 发展。然而，对早孕(FD)后胎儿死亡的遗传学研究，DDS的极端结果， 到目前为止在范围和规模上都受到限制。我的导师建立了一个国际胎儿组织 基因组学联盟将对10,500个样本(3,500个FD病例及其家庭成员)进行测序 疑似遗传病因的FD现在提供了一个前所未有的机会来阐明最极端的 对人类至关重要的单个基因和剂量敏感基因组片段的突变后果 发展。在这个团契中，我将整合来自这个队列的全基因组测序和尸检数据 以及来自先前建立的DD队列的数据，以发现突变并从功能上表征突变 人类基因组中不耐受的基因座。我将首先定义跨约束指标的FD的遗传变异模式 (功能丧失、错义和非编码)和基因组变异类别(点突变、插入片段、结构突变 变异和重复扩增)，并研究这些模式中关于胎儿性别和突变的偏向 原籍父母。然后，我将调整一个能够整合证据的疾病协会统计框架 从所有编码和非编码变量类中，使用受约束区域的优先顺序，我将应用于 在FD中进行新的基因发现(目标1)。我将利用这些发现来生成以下功能预测 通过定义活动的生物网络和它们可能所在的细胞类型来定义突变不耐受的基因座 在开发早期运行(目标2)。最后，我将在不支持这些假设的DDS中测试这些函数假设 导致FD，包括活产胎儿结构异常、神经发育障碍和先天性 异常(目标3)。在这些研究目标的同时，一个由六名导师和顾问组成的非凡团队 多个学科、职业阶段和机构将提供教学培训、实践研究支持、 定期在研讨会和会议上发表演讲，以及各种软技能的发展 在我的博士培训期间，这些课程与我的职业目标直接一致。总而言之， 这项提案将利用独特的工具和资源，对 沿着发育异常和进化制约的连续体的极端，并将作为一个 在计算、统计和功能疾病基因组学方面为我提供了绝佳的培训机会。

Abstract Natural selection purges deleterious mutations from populations in genomic regions impacting survival or reproductive capacity. Recent genetic studies of massive population cohorts have revealed a continuous distribution across the human genome of this constraint on deleterious mutations. Large-scale association studies have found mutations in strongly constrained genomic regions to be major risk factors in many childhood developmental disorders (DDs), suggesting that highly constrained sequences are likely to play key roles in development. However, genetic studies of fetal demise after the first trimester (FD), an extreme outcome of DDs, have thus far been limited in scope and size. The establishment by my mentors of an international Fetal Genomics Consortium to sequence 10,500 samples (3,500 FD cases and their family members) ascertained for FD of suspected genetic etiology now offers an unprecedented opportunity to illuminate the most extreme consequences of mutation across individual genes and dosage sensitive genomic segments critical for human development. In this fellowship, I will integrate whole-genome sequencing and autopsy data from this cohort together with data from previously established DD cohorts to discover and functionally characterize mutationally intolerant loci in the human genome. I will first define patterns of genetic variation in FD across constraint metrics (loss-of-function, missense, and noncoding) and genomic variation classes (point mutations, indels, structural variants, and repeat expansions), and investigate biases in these patterns with respect to fetal sex and mutational parent-of-origin. I will then adapt a statistical framework for disease association capable of integrating evidence from all coding and noncoding variant classes with prioritization of constrained regions, which I will apply to perform novel gene discovery in FD (Aim 1). I will leverage these findings to generate functional predictions of mutationally intolerant loci by defining the biological networks of activity and the cell types in which they are likely to operate early in development (Aim 2). Finally, I will test these functional hypotheses across DDs that do not result in FD, including liveborn fetal structural abnormalities, neurodevelopmental disorders, and congenital anomalies (Aim 3). In parallel with these research aims, an exceptional team of six mentors and advisors across multiple disciplines, career stages, and institutions will provide didactic training, hands-on research support, regular opportunities for presentation in seminars and conferences, and a variety of soft skill development sessions that directly align with my career objectives during my PhD training. Collectively, the aims outlined in this proposal will take advantage of unique tools and resources to yield novel insights into the etiology of the extremes along the continuum of developmental anomalies and evolutionary constraint, and will serve as an outstanding training opportunity for me in computational, statistical, and functional disease genomics.