MAY 19, 2017
The latest version of BisQue integrates CellProfiler (http://cellprofiler.org/), a software for the quantitative analysis of biological images, developed at the Broad Institute. CellProfiler is designed for modular, flexible, high-throughput analysis of images, measuring size, shape, intensity, and texture of every cell (or other object) in one or more images. At the core of CellProfiler is the image analysis "pipeline", a series of steps that each perform an image processing function such as illumination correction, object identification (segmentation), and object measurement.
The new BisQue CellProfiler module allows to run arbitrary CellProfiler pipelines directly within BisQue. Pipelines are regular BisQue resources and can be viewed in the browser, annotated, and shared with others. The module can run any such pipeline on a single image or on a dataset of multiple images. BisQue's module execution system automatically parallelizes the execution on a compute cluster. Users can also run pipelines with parameter variations to study the effects of specific parameters. The results of a pipeline execution is stored back in BisQue as table and image resources linked to the module execution. Geometric entities stored in output tables (e.g., cell centroids) can automatically be overlayed as color-coded graphical objects with labels when viewing the image. Both generated tables and graphical objects can be processed further in additional BisQue modules or shared with other users.
Six example CellProfiler pipelines together with test images have already been uploaded into BisQue for convenience. These pipelines are: Human Cells, Fruit Fly Cells, Tumors, Comet Assay, Tissue Neighbors, and Wound Healing. Details on each of these examples can be found at http://cellprofiler.org/examples/.
A step-by-step walk-through for one of the pipelines can be found at https://biodev.ece.ucsb.edu/projects/bisquik/wiki/CellProfilerModule.