Time series expression patterns reveal the molecular processes of pancreatic cancer progression

Pancreatic cancer is a fatal malignant tumor with no obvious characteristics in the early stage of onset and high metastatic ability which results in a low survival. Understanding the detailed process of pancreatic cancer contributes to new treatments to prolong patients' survival.

We carried out an in-depth analysis by modularization while seeking for critical genes in the pathogenesis of pancreatic cancer so as to identify the molecular mechanisms of the condition using differential analysis, co-expression module analysis, enrichment analysis, and network connectivity analysis. In light of the hypergeometric test, ncRNA (non-coding RNA) and transcription factors that regulate the module would be predicted.

Conclusively, seven co-expression modules were obtained, in which CPA2 and A1BG were significantly differentially expressed in patients who had pancreatic cancer with active regulation in dysfunction modules. The modular genes significantly participated in second-messenger-mediated signaling as well as cellular calcium homeostasis and also controlled the interactions of neuroactive ligand-receptor. Besides, we identified ncRNA pivot including FENDRR and miR-92a-3p as well as transcription factors pivot including SPI1, STAT5A which significantly regulated the dysfunction module.

This study can help reveal core dysfunction modules, potential regulatory factors, and driver genes for pancreatic cancer, enhancing the understanding of its pathogenesis and providing a reference for prediction with respect to the survival time of patients with this condition.