Developing an interactive genome analytics tool

Scientists in Agriculture and Food at CSIRO are applying machine learning techniques to resolve biological factors and interactions underpinning variation in crop traits. We established a toolset capable of visualising the multi-layered output of these models. However, existent tools require unique data processing pipelines and manual manipulation of the data in order to produce interpretable results. We worked to develop a self-contained tool in Shiny, a visualisation framework in R, that incorporates JavaScript visualisation libraries as HTML widgets. This Genome Analytics tool enables CSIRO researchers to interactively explore their data in real-time and extract useful meaning in an accessible and automated way. We will describe our development approach, from the initial phase of exploration through to coding the visualisation framework. We will also discuss the considerations that were made to ensure we developed an application capable of informing scientific research.