Candidate silencer elements for the human and mouse genomes
Candidate silencer elements for the human and mouse genomes
复制标题
DOI：
10.1038/s41467-020-14853-5
复制
发表时间：
2020-02-26
影响因子：
16.6
通讯作者：
Hawkins, R. David
中科院分区：
文献类型：
--
作者：
Jayavelu, Naresh Doni;Jajodia, Ajay;Hawkins, R. David
基因调控的研究主要集中在控制基因激活或增加表达水平上。基因抑制或沉默的过程同样重要。染色质标签已经确定了增强子，然而，缺乏通过计算或实验方法对沉默子的全基因组鉴定。在这里，我们首先定义未表征的顺式调控元件可能含有沉默剂，并发现41.5%的类似7500测试元素显示沉默剂活性，使用大规模平行报告分析（MPRA）。我们基于MPRA数据训练了一个支持向量机分类器，以预测100多种人类和小鼠细胞或组织类型中的候选沉默剂。预测的候选消声器表现出预期的消声器的特性。利用启动子捕获HiC数据，我们发现超过50%的沉默子与表达非常低或没有表达的基因启动子相互作用。我们的研究结果提出了一个通用的策略，全基因组范围内的沉默元件的识别和表征。
The study of gene regulation is dominated by a focus on the control of gene activation or increase in the level of expression. Just as critical is the process of gene repression or silencing. Chromatin signatures have identified enhancers, however, genome-wide identification of silencers by computational or experimental approaches are lacking. Here, we first define uncharacterized cis-regulatory elements likely containing silencers and find that 41.5% of similar to 7500 tested elements show silencer activity using massively parallel reporter assay (MPRA). We trained a support vector machine classifier based on MPRA data to predict candidate silencers in over 100 human and mouse cell or tissue types. The predicted candidate silencers exhibit characteristics expected of silencers. Leveraging promoter-capture HiC data, we find that over 50% of silencers are interacting with gene promoters having very low to no expression. Our results suggest a general strategy for genome-wide identification and characterization of silencer elements.