Across the spectrum of cancer types, pancreatic cancer has one of the worst prognoses[1, 2]. Fewer than 10% of patients diagnosed with pancreatic cancer survive more than a year after diagnosis . The research surrounding pancreatic cancer has predominantly focused on the mutations associated with the disease. In contrast, the changes in expression associated with pancreatic cancer have not been studied to the same depth. Studying the transcriptional changes associated with specific types of cancer has greatly benefited to our understanding of the disease and has identified subtypes of cancers with distinct prognostic signatures that are histologically indistinguishable. These studies have facilitated the development of accurate methods to diagnose the disease as well as therapeutic treatments tailored to the specific molecular changes associated with each subtype of cancer [3-10]. Here, we have used microarrays to characterise gene expression from primary pancreatic cancer samples from 117 individuals. This information was used to identify the genes differentially expressed between tumourous and normal tissue, predict the molecular subtype of each sample and confirm the prognostic importance of the different subtypes of pancreatic ductal adenocarcinoma. The methods used to characterise the primary tumour samples were applied to a panel of cell lines and identified the first cell lines belonging to the exocrine-like subtype. To capture the changes in expression not described by microarrays analysis, this work was supplemented by RNA-Seq. This technique allowed for a high resolution, study of expression in the tumorous and normal pancreatic tissue, from two patients. The single nucleotide resolution of this data made it possible to characterise expression at the level of genes and individual transcripts. This information was used to define differential gene expression, identify alternative transcript usage, and describe novel transcriptional events arising from known pancreatic cancer genes.