A Genetic Algorithm for Pancreatic Cancer Diagnosis.

Pancreatic cancer is one of the leading causes of cancer-related death in the industrialized countries and it has the least favorable prognosis among various cancer types. In this study we aim to facilitate early detection of the pancreatic cancer by finding minimal set of genetic biomarkers that can be used for establishing diagnosis. We propose a genetic algorithm and we test it on gene expression data of 36 pancreatic ductal adenocarcinoma tumors and matching normal pancreatic tissue samples. Our results show that a minimum group of genes are able to constitute a high reliability pancreatic cancer predictor.