Duration: 35:11

131 views

1

Phaedra is an Eclipse-based application designed for supporting the High Content Screening (HCS) work flow. HCS is a life science research method that offers several advantages over more traditional screening methods, and has been instrumental in providing new insights into pharmaceutical research data. However, it also poses some technical challenges, such as a greatly increased data volume, and the requirement to navigate and process that data in real-time.

Phaedra aims to provide an open source platform that tackles those challenges, while offering users a rich, interactive experience based on the Eclipse platform.
Using the familiar concept of an Eclipse workbench, Phaedra adds dozens of views that respond to each other's selections, form editors with input validation, an extensive and deeply configurable charting framework, scripting integration for Javascript, Python, R, Matlab and more.

Since HCS is all about digital imaging, and those images are large and numerous, Phaedra uses the JPEG2000 technology to compress, store and render those images efficiently. Compression factors are significantly higher than in traditional jpeg or gzip compressions, and the ability is preserved to render images - or regions of images - at any size, directly from the compressed file. The image viewers that are included in Phaedra offer real-time zoomable and scrollable renderers, again with full interactivity.
This allows a user, for example, to select a group of cells on an image, and immediately plot the properties of those cells in a 1D, 2D or 3D chart.

In this talk we will present the functionality of Phaedra and demonstrate the typical workflow of analyzing high-content screening data. Since many components of Phaedra may be of broader interest than HCS alone, we hope to trigger discussions with the Eclipse Science Working group and the Eclipse science community more generally to strengthen the position of Eclipse in the life sciences area.

Speaker(s): Frederick Michielssen [Open Analytics NV], Tobias Verbeke [Open Analytics NV]