生物学杂志

生物学杂志 ›› 2021, Vol. 38 ›› Issue (2): 105-.doi: 10.3969/j.issn.2095-1736.2021.02.105

• 技术方法 • 上一篇    下一篇

基于绝缘密度HiCID算法的染色质接触域边界检测

  

  1. 云南民族大学电气信息工程学院，昆明650504
  • 出版日期:2021-04-18 发布日期:2021-04-25
  • 通讯作者: 丰继华,博士,副教授,研究领域为生物信息学, E⁃mail:jihuafeng@yahoo.com
  • 作者简介:黄月月,硕士研究生,研究方向为生物信息处理, E⁃mail:1276127902@ qq.com
  • 基金资助:
    国家自然科学基金(31160234);云南民族大学研究生创新基金项目(2018YJCXS185)

The identification of chromatin interaction domains based on HiCID

  1. School of Electrical and Information Engineering，Yunnan University for Nationalities，Kunming 650504，China
  • Online:2021-04-18 Published:2021-04-25

PDF (PC)

103

摘要： 针对现有基于Hi⁃C数据的接触域边界检测算法进行改进，提出用绝缘密度统计量来表征接触域边界强度变化的绝缘密度算法（Hi⁃C insulation density，HiCID）。HiCID算法将网络增强技术嵌入预处理过程对原始Hi⁃C数据进行降噪，根据绝缘子结合蛋白（CTCF）与组蛋白修饰的富集丰度确定域边界筛选阈值，同时为不同分辨率Hi⁃C数据优化选择滑动窗口的大小和数量。实验结果表明，提出的接触域边界检测算法在一致性和准确性上均优于HiCDB算法，具有不依赖于特定物种的可移植性特点，且算法随Hi⁃C数据分辨率提高表现出更好的性能。

关键词: 染色质接触域, 定向分化, 网络增强, 绝缘密度算法, 绝缘子结合蛋白

Abstract: This paper improved the existing contact domain boundary detection algorithm based on Hi⁃C data，and proposed HiCID（Hi⁃C insulation density）algorithm which used insulation density statistics to characterize the strength change of the contact domain boundary. The HiCID algorithm embeded network enhancement technology into the pre⁃processing process to denoise the raw Hi⁃C data. Based on the enrichment abundance of insulator binding protein（CTCF）and histone modification，the domain boundary screening threshold was determined. At the same time，the right size and number of sliding windows for Hi⁃C data with different resolutions could be optimally selected. The experimental results showed that the contact domain boundary detection algorithm proposed was superior to the HiCDB algorithm in terms of consistency and accuracy，and had the characteristics of portability that did not depend on specific species，and the algorithm showed better performance as Hi⁃C data resolution increases.

Key words: chromatin contact domain, lineage commitment, network enhancement, Hi?C insulation density, CTCF

中图分类号: