Enabling Cloud and Grid Powered Image Phenotyping

Ability to effectively manage high resolution and high throughput image data is a universal need for phenotyping related analysis. With recent advances in automation, robotics, super resolution optics etc. the size and rate of data generation is fast outpacing traditional methods for analysis and data lifecycle management.

Built on iPlant’s cyberinfrastructure and in collaboration with the Bio-Image informatics center (University of California Santa Barbara) and Edgar Spalding (University of Wisconsin Madison) a scalable, web based, high throughput image analysis and management platform BISQUE (Bio-Image Semantic Query User Environment) is available for researchers.

This standards compliant platform provides access to integrated high throughput data transfers and scalable compute environment, with its rich web interface that allows seamless access for managing high resolution (20Kx20K) images using tiling techniques. Incorporating new analysis methods to leverage this versatile and flexible platform is easy, allowing computational groups to make their analysis methods available with web interface. BISQUE can extract metadata (such as latitude and longitude) and overlay data on maps.

For image analysis routines with custom GUI (Graphical User Interface) that are not suitable for web interfaces, access to iPlant’s cloud infrastructure (Atmosphere) and data store is available.

Combination of the these technologies allows researchers to access and utilize this infrastructure right from their desktops, work collaboratively in groups and sharing large files and resultant analysis.