Identifying biomarker candidates of influenza infection based on scalable time-course big data of gene expression

Large-scale biological data processing is an important topic in computational biology, which gives the promotion of many biomedical research. For example, a huge volume of gene expression data can help us achieve potentially useful medical knowledge and identify disease biomarker candidates. Nevertheless, a great deal of attention has been paid to unscalable single-time-point expression data after disease symptoms appear (outbreak period) and there have been few investigations into scalable time-course expression data before disease symptoms appear (incubation) for each sample. By exploiting such dynamic big data in the incubation, we can easily catch early signals of disease states and prevent illness in the first place. In this study, we apply a new mathematical model on biological data of given incubations and identify biomarker candidates using an intellectualized method. The model narrows a large number of alternative genes into a few ones (top genes), which facilitate the discovery of genes related to disease. The aim of our work is to propose a powerful biomarker-detecting tool, which helps people to aid early diagnosis or identify effective drug targets before the appearance of clinical symptoms.