Min Xiao-Shan, Huang Peng, Liu Xu, Dong Chao, Jiang Xiao-Lin, Yuan Zheng-Tai, Mao Lin-Feng, Chang Shi
Department of Ophthalmology, Xiangya Hospital of Central South University, Changsha, Hunan, 410008, Peoples Republic of China.
Department of General Surgery, Xiangya Hospital of Central South University, #87 Xiangya Road, Changsha, Hunan, 410008, Peoples Republic of China.
Tumour Biol. 2015 Sep;36(10):7457-63. doi: 10.1007/s13277-015-3410-6. Epub 2015 Apr 24.
This study was aimed to identify the prognostic risk markers for thyroid papillary carcinoma (TPC) by bioinformatics. The clinical data of TPC and their microRNAs (miRNAs) and genes expression profile data were downloaded from The Cancer Genome Atlas. Elastic net-Cox's proportional regression hazards model (EN-COX) was used to identify the prognostic associated factors. The receiver operating characteristic (ROC) curve and Kaplan-Meier (KM) curve were used to screen the significant prognostic risk miRNA and genes. Then, the target genes of the obtained miRNAs were predicted followed by function prediction. Finally, the significant risk genes were performed literature mining and function analysis. Total 1046 miRNAs and 20531 genes in 484 cases samples were identified after data preprocessing. From the EN-COX model, 30 prognostic risk factors were obtained. Based on the 30 risk factors, 3 miRNAs and 11 genes were identified from the ROC and KM curves. The target genes of miRNA-342 such as B-cell CLL/lymphoma 2 (BCL2) were mainly enriched in the biological process related to cellular metabolic process and Disease Ontology terms of lymphoma. The target genes of miRNA-93 were mainly enriched in the pathway of G1 phase. Among the 11 prognostic risk genes, v-maf avian musculoaponeurotic fibrosarcoma oncogene homologue F (MAFF), SRY (sex-determining region Y)-box 4 (SOX4), and retinoic acid receptor, alpha (RARA) encoded transcription factors. Besides, RARA was enriched in four pathways. These prognostic markers such as miRNA-93, miRNA-342, RARA, MAFF, SOX4, and BCL2 may be used as targets for TPC chemoprevention.
本研究旨在通过生物信息学方法鉴定甲状腺乳头状癌（TPC）的预后风险标志物。从癌症基因组图谱下载了TPC的临床数据及其微小RNA（miRNA）和基因表达谱数据。使用弹性网Cox比例回归风险模型（EN-COX）来鉴定预后相关因素。采用受试者工作特征（ROC）曲线和Kaplan-Meier（KM）曲线筛选具有显著预后风险的miRNA和基因。然后，预测所得miRNA的靶基因并进行功能预测。最后，对显著风险基因进行文献挖掘和功能分析。经过数据预处理，在484例样本中鉴定出总共1046个miRNA和20531个基因。从EN-COX模型中获得了30个预后风险因素。基于这30个风险因素，从ROC曲线和KM曲线中鉴定出3个miRNA和11个基因。miRNA-342的靶基因如B细胞淋巴瘤/白血病-2（BCL2）主要富集在与细胞代谢过程相关的生物学过程以及淋巴瘤的疾病本体术语中。miRNA-93的靶基因主要富集在G1期途径中。在这11个预后风险基因中，v-maf禽肌纤维肉瘤癌基因同源物F（MAFF）、SRY（性别决定区Y）-盒4（SOX4）和视黄酸受体α（RARA）编码转录因子。此外，RARA富集在四条途径中。这些预后标志物如miRNA-93、miRNA-342、RARA、MAFF、SOX4和BCL2可作为TPC化学预防的靶点。
