基于单细胞多组学数据解析结直肠癌动态调控扰动  

Analysis of regulatory perturbations of malignant transformation in colorectal cancer based on single-cell multi-omics data

作　　者：徐琪[1] 禹雷 潘涛 张娅 李永生 XU Qi;YU Lei;PAN Tao;ZHANG Ya;LI Yong-sheng(College of Biomedical Information and Engineering,Hainan Medical University,Haikou 571199,Hainan,CHINA)

机构地区：[1]海南医学院生物医学信息与工程学院,海南海口571199

出　　处：《海南医学》2024年第11期1533-1544,共12页Hainan Medical Journal

基　　金：海南医学院国家级自然科学研究项目培育基金重点项目(编号:JBGS202103);海南省自然科学基金青年基金项目(编号:823QN249)。

摘　　要：目的探讨结直肠癌恶性转变过程中的分子特征及其调控网络扰动,深入剖析结直肠癌恶性转变的微环境异质性并识别潜在治疗靶点。方法首先从Gene Expression Omnibus(GEO)数据库中收集单细胞转录组测序数据,从欧洲分子生物学实验室(EMBL)收集单细胞ATAC测序数据。数据共涉及来自14例患者的70例样本,包含22例正常样本、43例息肉样本和5例肿瘤样本。首先对scRNA-Seq测序数据进行预处理,基于Seurat包筛除双细胞和低质量细胞,对scATAC-Seq测序数据使用10X Cell Ranger-atac进行原始数据处理,并使用Signac包去除低质量细胞,得到三个阶段的scRNA-Seq和scATAC-Seq细胞图谱。进一步数据整合,基于Pando R包推断T细胞和上皮细胞的转录调控网络,通过mfinder软件分析网络motif和拓扑属性,对转录因子进行分级分析。最后,基于随机森林算法进行细胞分子特征预测及预后分析。结果筛选出202465个scRNA-Seq测序细胞和136422个scATAC-Seq测序细胞。基于数据整合分析,构建不同癌变阶段的细胞图谱。结果发现,随着结直肠癌进展,肿瘤微环境组成发生显著变化,尤其是T细胞和上皮细胞在不同疾病阶段的比例有较大差异。基于单细胞多组学数据整合,利用Pando包推断结直肠癌不同阶段的转录调控网络,揭示了T细胞和上皮细胞中转录因子及其调控关系的动态变化。功能富集分析结果显示在T细胞和上皮细胞中,转录因子所调控的功能在不同的疾病阶段有明显差异。基于转录调控网络分析发现,T细胞和上皮细胞转录调控网络符合无标度网络特性。网络motif分析揭示了在不同阶段存在的特定motif模式,反映了网络拓扑结构的动态变化,且网络中大多数相互作用都具有阶段特异性。共享转录因子的层级在癌变过程中也会发生变化。最后基于转录因子调控网络构建的分类器可以成功识别T细胞和上皮细胞,表明其作�Objective To explore the molecular characteristics and regulatory network perturbations during the malignant transformation process of colorectal cancer,and to deeply analyze the heterogeneity of microenvironment and potential therapeutic targets of malignant transformation of colorectal cancer.Methods We collected single-cell transcriptome sequencing from the Gene Expression Omnibus(GEO)database and single-cell ATAC sequencing data from the European Molecular Biology Laboratory(EMBL).The data encompassed 70 samples from 14 patients,including 22 normal samples,43 polyp samples,and 5 tumor samples.Initially,preprocessing of scRNA-Seq sequencing data involved the removal of double cells and low-quality cells based on the Seurat package.For scATAC-Seq sequencing data,the 10X Cell Ranger-atac was used for initial data processing,and the Signac package was employed to eliminate low-quality cells,resulting in scRNA-Seq and scATAC-Seq maps for three stages.Subsequently,data integration was performed,and the Pando R package was utilized to infer the transcriptional regulatory networks of T cells and epithelial cells.Network motif and topological attribute analysis,as well as hierarchical analysis of transcription factors,were conducted using the mfinder software.Finally,cell molecular feature prediction and prognosis analysis were carried out using the random forest algorithm.Results A total of 202465 scRNA-Seq sequencing cells and 136422 scATAC-Seq sequencing cells were screened out.Through data integration analysis,cell atlases were constructed for different stages of carcinogenesis.The results revealed significant changes in the tumor microenvironment as colorectal cancer progresses,particularly in the substantial differences in the proportions of T cells and epithelial cells at different disease stages.Utilizing integrated single-cell multi-omics data and the Pando package,the study infered the transcriptional regulatory networks in different stages of colorectal cancer,elucidating the dynamic changes in transcriptio

关 键 词：结直肠癌 单细胞多组学数据 转录调控网络 细胞分子特征 

分 类 号：R735[医药卫生—肿瘤]