Inferring diploid 3D chromatin structures from bulk and single-cell Hi-Cdata

项目总结/摘要 基因组的3D组织在许多细胞过程中起着关键作用，例如基因调控， 分化和细胞周期。像Hi-C这样的测定以高通量的方式测量DNA-DNA接触。 从这些数据中推断染色体如何折叠的准确3D模型可以产生隐藏在染色体中的洞察力。 原始数据存在许多方法来推断单倍体基因组的3D结构，但二倍体基因组对染色体的三维结构的推断存在很大困难。 这是一个更具挑战性的问题，因为Hi-C数据本身并不能区分等位基因。此外，本发明还 虽然单细胞实验已经清楚地表明，染色质结构表现出很大的异质性， 在一个群体中，单细胞Hi-C数据的稀疏性给推断带来了额外的困难。我们有 最近发表了一种方法，通过建立在我们以前的概率框架上来推断3D二倍体基因组， 为单倍体数据而开发。我们建议将这种方法应用于二倍体酵母基因组的建模，以进一步 在酵母中表征有丝分裂同源配对。我们还建议将这种方法扩展到单细胞 数据，并验证和整合我们的方法与染色质位点和核蛋白的显微镜。我们将 从而提供高分辨率成像和DNA序列的集成3D模型。

Project Summary/Abstract The 3D organization of the genome plays a key role in many cellular processes, such as gene regulation, differentiation, and the cell cycle. Assays like Hi-C measure DNA-DNA contacts in a high-throughput fashion. Inferring from such data accurate 3D models of how chromosomes fold can yield insights that are hidden in the raw data. Many methods exist to infer the 3D structures of haploid genomes, but diploid genomes pose a much more challenging problem because Hi-C data does not inherently distinguish between the alleles. Additionally, while single-cell experiments have made clear that chromatin structure exhibits a great deal of heterogeneity within a population, the sparsity of single-cell Hi-C data poses additional difficulties for inference. We have recently published a method to infer 3D diploid genomes by building upon a probabilistic framework we previously developed for haploid data. We propose to apply this method to model diploid yeast genomes in order to further characterize mitotic homolog pairing in yeast. We also propose to extend this method to work with single-cell data, and to validate and integrate our method with microscopy of chromatin sites and nuclear proteins. We will thereby provide an integrated 3D model of high-resolution imaging and DNA sequence.