The advances in DNA-sequencing technology led to a vast amount of datasets from diverse biological sources. Making those complex datasets accessible for data mining to non-bioinformaticians in a timely manner is still challenging. In a typical scenario unprocessed data is generated, pre-processed and distributed by a third party (e.g. a DNA-sequencing center) before biological questions are addressed. While R offers excellent capabilities to do that, its steep learning curve makes it rather unattractive for pure biologists, who “fear the command line” and prefer graphical user interfaces.
Using shiny apps we present examples of how to make biological data (gene expression datasets) and their associated analyses methods available to biologists. Data mining in gene expression datasets can involve techniques like dimensional reduction, clustering, testing for significant differences or term enrichment analyses. In addition to conducting those analyses, data visualization is necessary to quickly evaluate results. Those tasks normally require at least a working knowledge of R and some commonly used packages.
The combination of shiny apps with common analyses packages allows the rapid deployment of applications tailored to a specific dataset that allows effective data mining with a graphical user interface. By including raw data and pre-calculated results objects shiny apps can be used as an efficient, self-contained way of data distribution and analysis.