Direct Determination of Multiple Specific Forms of DNA Chemical Modifications in Human Genome

项目总结/摘要 DNA的信息内容不限于一级序列（A，C，G，T），还可以通过 个别碱基的化学修饰。例如，DNA甲基化，特别是5-甲基胞嘧啶（5 mC）， 由于其在人类发育和疾病中的重要调节作用而被广泛研究。此外该 发现由泰特酶介导的5 mC主动去甲基化为5-羟甲基胞嘧啶（5 hmC）， 5-甲酰胞嘧啶（5 fC）和5-羧基胞嘧啶（5caC）揭示了对细胞生长动力学性质的深刻见解。 人类甲基化及其与多种人类疾病的密切相关性。除了这些化学修饰 我们和其他人最近的研究发现，N6-甲基腺嘌呤（6 mA），另一种形式的胞嘧啶， 甲基化以前被认为只存在于细菌和原生动物中，也存在于真核生物基因组中 包括人类基因组。除了这些表观遗传标记，不同形式的DNA损伤代表了 另一类具有重要生物学意义的DNA化学修饰。虽然少数 已经开发了用于绘制单个化学修饰的方法，并且其中一些方法被广泛使用， 对于广泛的研究人员来说，通常很难掌握每一种协议来映射每一种形式的修改。第三， 一代测序技术支持DNA修饰的直接检测，他们面临的基本问题是， 区分不同形式的修改的挑战。该项目的目标是开发一个 用于直接绘制多种形式的DNA甲基化和DNA损伤事件的新技术 同步其核心思想是，每种形式的核酸修饰都有一个独特的签名， 它们与DNA聚合酶或第三代测序中的纳米孔的物理相互作用; 可以通过深度学习方法对签名进行建模。我们将开发这项技术， 创新战略，以应对一些基本挑战，然后全面评估 技术，以促进更广泛的应用。

PROJECT SUMMARY/ABSTRACT The information content of DNA is not limited to the primary sequence (A, C, G, T), but is also conveyed by chemical modifications of individual bases. For example, DNA methylation, specifically 5-methylcytosine (5mC), has been widely studied for its important regulatory roles in human development and diseases. In addition, the discovery of active demethylation of 5mC, mediated by TET enzymes, into 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC) and 5-carboxylcytosine (5caC) revealed great insights into the dynamic nature of the human methylome and its close relevance to multiple human diseases. Beyond these chemical modifications to cytosine, recent studies by us and others discovered that N6-methyladenine (6mA), another form of methylation previously thought exclusively existing in bacteria and protozoa, also exists in eukaryotic genomes including the human genome. In addition to these epigenetic marks, different forms of DNA damages represent another category of DNA chemical modifications that are of important biological relevance. Although a few methods for mapping individual chemical modifications have been developed and some are widely used, it is usually hard for broad researchers to master every protocol to map each form of modification. While third- generation sequencing technologies support the direct detection of DNA modifications, they face fundamental challenges distinguishing among different forms of modifications. The objective of this project is to develop a novel technology for the direct mapping of multiple forms of DNA methylation and DNA damage events simultaneously. The core idea is that each form of nucleic acid modification has a unique signature in terms of their physical interaction with DNA polymerase, or nanopores in third-generation sequencing; and these signatures can be modeled by deep learning methods. We will develop this technology using multiple innovative strategies to address a few fundamental challenges, and then comprehensively evaluate the technology to facilitate broad applications.