胰腺癌增生分化及凋亡相关蛋白互作动态网络特征分析

Analysis of protein interaction networks related to hyperplasia differentiation and apoptosis of pancreatic cancer

目的 利用生物信息学技术查找与建立胰腺癌相关基因、单核苷酸多态性（SNP）、微小RNA（miRNA）、蛋白质调控网络以及蛋白通路（pathway）功能注释等信息数据,并建立胰腺癌增生分化及凋亡相关蛋白互作动态网络.方法 通过文本挖掘技术和已有胰腺癌数据库（PC-GDB、Ensembl）进行数据收集和整理,然后利用倍数变化（FC）方法对高通量基因表达数据库（GEO）中胰腺癌相关基因表达谱芯片进行分析,计算正常组和疾病组的差异表达基因,得到和胰腺癌相关的数据信息.进一步将GEO数据库中获取的3组基因表达谱芯片（编号：GSE22780、GSE22973、GSE14245）筛选后,将基因芯片分为2组（正常组、胰腺癌组）,利用FC方法得到胰腺癌的异常表达基因集,再将得到的基因集投射到蛋白质相互作用关系,得到相应的蛋白质网络,并取其中相对独立、相互作用较为集中且与增生分化及凋亡密切相关的子网络进行功能注释富集分析.结果 GEO数据库中得到差异基因有1 766个,PC-GDB和Ensembl数据库中潜在与胰腺癌相关基因1 173个,胰腺癌相关miRNA共140个,疾病相关SNPs共501个.建立了胰腺癌发生发展过程中的蛋白质相互作用动态网络,并分析其中主要发挥调节增生、诱导分化及凋亡作用的子网络.结论 利用文本挖掘及生物信息学技术收集了胰腺癌相关基因和蛋白数据,并建立了胰腺癌增生分化及凋亡相关蛋白互作动态网络,该网络可为后续针对胰腺癌的研究提供线索和支持,尤其可用于寻找胰腺癌的诊治靶点.

Objective Using bioinformatics methods to analyze large amounts of data generated by pancreatic cancer gene chips in order to establish protein interaction networks related to hyperplasia differentiation and apoptosis of pancreatic cancer.Methods Differentially expressed genes between pancreatic cancer and normal pancreas were found by the method of text mining and fold change （FC）.Then use FC to analyses 3 chips （NO.GSE22780, GSE22973, GSE14245） form Gene Expression Omnibus （GEO）database in order to find differentially expressed genes during pancreatic cancer phase.Furthermore, the differentially expressed genes were projected on the protein interaction relations in order to establish protein interaction networks.Finally functional annotation bioinformatics microarray analysis was performed to those relatively independent networks which were highly related to hyperplasia differentiation and apoptosis.Results 1 766 differentially expressed genes were found in GEO database and 1 173 differentially expressed genes were found in PC-GDB and Ensembl database.140 microRNAs and 501 single nucleotide polymorphisms were found related to pancreatic cancer.Protein interaction networks related to hyperplasia differentiation and apoptosis of pancreatic cancer was established.Conclusion This study has collected genes and protein data related to pancreatic cancer by using bioinformatics methods.Established protein interaction networks related to hyperplasia differentiation and apoptosis of pancreatic cancer, which can provide new leads to following researches.