非典型DNA结构在基因转录、DNA复制、基因组突变等生物学事件中发挥重要功能，然而，由于已有预测方法的局限性和组学数据积累有限，对哺乳动物染色质中非典型DNA结构位点组织特异性的研究仍非常初步。已有研究发现非典型DNA结构位点倾向于具有特殊的核小体定位特征，考虑到在人类和小鼠的几十种细胞类型中都有已发表的核小体定位数据，因此通过整合核小体定位特征有望解析非典型DNA结构位点的组织特异性；然而，迄今国际上还没有相关的报道。在前期的研究中，申请人发展了一系列核小体定位数据分析方法，并揭示了核小体定位模式的建立规律。在此基础上本项目提出，基于已发表的人类和小鼠的MNase-seq数据，拟构建整合核小体定位特征的非典型DNA结构位点预测模型，并应用该模型解析哺乳动物染色质中非典型DNA结构位点的组织特异性及其作用。本项目所提出的核小体定位数据的分析策略可以推广到研究染色质上的其它生物学事件。

Non-canonical DNA structures are involved in various biological processes including gene transcription, DNA replication and mutagenesis with essential biological functions. However, the tissue specificity of non-canonical DNA structures in mammalian genome is still poorly understood, mainly limited by the unsatisfied performance of computational approaches and the lack of omics data in various tissues. The relationships between non-canonical DNA structures and nucleosome organization was presented, and nucleosome organization profiles are available in dozens of human and mouse cell types, respectively. Therefore, it is possible to illuminate the tissue specificity of non-canonical DNA structures through integrative analysis of nucleosome organization features. However, no related studies have been published. In previous studies, we developed a series of computational methods in analyzing nucleosome organization data, and revealed the principles of nucleosome organization establishment. On the basis of our previous studies, this project proposes to build logistic regression models with the integration of published MNase-seq data in human and mouse for predicting non-canonical DNA structures across the genome. By applying the prediction models, we aim to reveal the tissue specificity and biological functions of non-canonical DNA structures in mammalian genome. The strategy of nucleosome organization dataset analysis proposed by this project could also be applied in researches on other biological processes related with chromatin.