基于基因互作网络熵量化细胞分化状态
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

国家自然科学基金(11831015,91730301)


Quantifying the state of cell differentiation based on the gene networks entropy
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    细胞动态过程的研究表明,细胞在动态过程中会发生状态变化,主要由细胞内部的基因表达情况控制。随着高通量测序技术的发展,大量的基因表达数据能够在单细胞水平上获得细胞真实的基因表达信息。然而,现有大多数研究方法需要使用除基因表达以外其他的信息,带来了额外的复杂度和不确定性。此外,普遍存在的“缺失值”事件更是影响了对细胞动态发展的研究。为此,文中提出了基因互作网络熵方法,来量化细胞的分化状态,以此来研究细胞的发展动态。具体而言,通过借助网络的稳定性,依据基因间的关联来构造细胞特异性网络,并定义基因互作网络熵,从而将不稳定的基因表达数据转换成相对稳定的基因互作网络熵。该方法没有额外的复杂度和不确定性,同时在一定程度上规避了“缺失值”事件的影响,能更可靠地表征诸如细胞命运等生物学过程。将该方法应用于头颈部鳞状细胞癌和慢性髓细胞白血病两个单细胞RNA-seq数据集,不仅能够有效区分恶性细胞与良性细胞、有效区分不同分化时期,还可以有效反映疾病疗效过程,证明了利用基因互作网络熵方法研究细胞动态的潜力。该方法旨在探索单细胞混乱程度水平的动态信息,从而研究生物系统进程中的动态变化情况。研究结果能够为细胞分化、追踪癌症发展以及疾病对药物反应过程等研究提供科学建议。

    Abstract:

    Studies of cellular dynamic processes have shown that cells undergo state changes during dynamic processes, controlled mainly by the expression of genes within the cell. With the development of high-throughput sequencing technologies, the availability of large amounts of gene expression data enables the acquisition of true gene expression information of cells at the single-cell level. However, most existing research methods require the use of information beyond gene expression, thus introducing additional complexity and uncertainty. In addition, the prevalence of dropout events hampers the study of cellular dynamics. To this end, we propose an approach named gene interaction network entropy (GINE) to quantify the state of cell differentiation as a means of studying cellular dynamics. Specifically, by constructing a cell-specific network based on the association between genes through the stability of the network, and defining the GINE, the unstable gene expression data is converted into a relatively stable GINE. This method has no additional complexity or uncertainty, and at the same time circumvents the effects of dropout events to a certain extent, allowing for a more reliable characterization of biological processes such as cell fate. This method was applied to study two single-cell RNA-seq datasets, head and neck squamous cell carcinoma and chronic myeloid leukaemia. The GINE method not only effectively distinguishes malignant cells from benign cells and differentiates between different periods of differentiation, but also effectively reflects the disease efficacy process, demonstrating the potential of using GINE to study cellular dynamics. The method aims to explore the dynamic information at the level of single cell disorganization and thus to study the dynamics of biological system processes. The results of this study may provide scientific recommendations for research on cell differentiation, tracking cancer development, and the process of disease response to drugs.

    参考文献
    相似文献
    引证文献
引用本文

关天昊,高洁. 基于基因互作网络熵量化细胞分化状态[J]. 生物工程学报, 2022, 38(2): 820-830

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2021-02-11
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2022-02-25
  • 出版日期:
您是第位访问者
生物工程学报 ® 2024 版权所有

通信地址:中国科学院微生物研究所    邮编:100101

电话:010-64807509   E-mail:cjb@im.ac.cn

技术支持:北京勤云科技发展有限公司