Identification of target gene and prognostic evaluation for lung adenocarcinoma using gene expression meta-analysis, network analysis and neural network algorithms

Gurudeeban Selvaraj, Satyavani Kaliamurthi, Aman Chandra Kaushik, Abbas Khan, Yong Kai Wei, William C. Cho, Keren Gu, Dong Qing Wei

Research output: Contribution to journalArticlepeer-review

45 Scopus citations

Abstract

Background: Lung adenocarcinoma (LUAD) is a heterogeneous disease with poor survival in the advanced stage and a high incidence rate in the world. Novel drug targets are urgently required to improve patient treatment. Therefore, we aimed to identify therapeutic targets for LUAD based on protein-protein and protein-drug interaction network analysis with neural network algorithms using mRNA expression profiles. Results: A comprehensive meta-analysis of selective non-small cell lung cancer (NSCLC) mRNA expression profile datasets from Gene Expression Omnibus were used to identify potential biomarkers and the molecular mechanisms related to the prognosis of NSCLC patients. Using the Network Analyst tool, based on combined effect size (ES) methods, we recognized 6566 differentially expressed genes (DEGs), which included 3036 downregulated and 3530 upregulated genes linked to NSCLC patient survival. ClueGO, a Cytoscape plugin, was exploited to complete the function and pathway enrichment analysis, which disclosed “regulated exocytosis”, “purine nucleotide binding”, “pathways in cancer”, and “cell cycle” between exceptionally supplemented terms. Enrichr, a web tool examination, demonstrated “early growth response protein 1 (EGR-1)”, “hepatocyte nuclear factor 4α (HNF4A)”, “mitogen-activated protein kinase 14 (MAP3K14)”, and “cyclin-dependent kinase 1 (CDK1)” to be among the most prevalent TFs and kinases associated with NSCLC. Our meta-analysis identified that MAPK1 and aurora kinase (AURKA) are the most obvious class of hub nodes. Furthermore, protein-drug interaction network and neural network algorithms identified candidate drugs such as phosphothreonine and 4-(4-methylpiperazin-1-yl)-n-[5-(2-thienylacetyl)-1,5-dihydropyrrolo[3,4-c]pyrazol-3-yl] benzamide and for the targets MAPK1 and AURKA, respectively. Conclusion: Our study has identified novel candidate biomarkers, pathways, transcription factors (TFs), and kinases associated with NSCLC prognosis, as well as drug candidates, which may assist treatment strategy for NSCLC patients.

Original languageEnglish
Pages (from-to)120-134
Number of pages15
JournalJournal of Biomedical Informatics
Volume86
DOIs
StatePublished - 1 Oct 2018
Externally publishedYes

Keywords

  • Hub nodes
  • Lung adenocarcinoma
  • Microarray data
  • Network analyst
  • Neural network
  • STRING
  • Walktrap module

ASJC Scopus subject areas

  • Health Informatics
  • Computer Science Applications

Fingerprint

Dive into the research topics of 'Identification of target gene and prognostic evaluation for lung adenocarcinoma using gene expression meta-analysis, network analysis and neural network algorithms'. Together they form a unique fingerprint.

Cite this