HiCSeg：針對不同樣本和物種的互動式基因體分割

Abstract

Hi-C的全基因組染色體接觸可用於研究染色體的更高級別組織，例如隔室或拓撲關聯域。根據哺乳動物Hi-C圖的主成分分析可得到數據中兩個區室A和B。TAD或隔室可被視為基因組的分段。通常我們會使用基因體分割進行數據壓縮，並在不同細胞類型中整理出不同的修飾。我們比較了不同解析度下的PCA結果以找出差異，然後引入ChIP-Seq數據進行進一步分析。我們還引進了其他兩種進行聚類的方法，Louvain和Leiden。它們不僅可以與PCA的結果進行比較，還可以計算出網路的相關性。此外，我們可以基於結合ChIP-Seq和Hi-C的資訊使用兩者相加及網路融合來分割基因組。

The genome-wide chromosomal contact by Hi-C can be used to investigate the higher-level organization of chromosomes, such as compartments or topologically associating domains (TAD). Hi-C data revealed two compartments, A and B, based on principal component analysis (PCA) of Hi-C maps in mammals. TAD or compartment can be considered as a segmentation of the genome. Generally, we use genome segmentation for data compression and sort out different modifications in different cell types. We compared the PCA results in various resolutions to determine the difference and introduced the ChIP-Seq data for further analysis. We also introduce other methods to do clustering, which are the Louvain and Leiden methods. They can not only compare with the result of PCA but also figure out the correlation of networks. Furthermore, we can segment the genome based on integrated ChIP-Seq and Hi-C information using adding function and network fusion.