Automatic detection of morphological motifs

Driscoll et al. published an article in the October issue of Nature Methods describing u-shape3D, a suite of computational tools to investigate how 3D cellular morphology governs intracellular signaling. In this article, we introduce a generic morphological motif detector that uses machine learning to find morphological structures, such as lamillipodia, blebs, and filopodia, given user provided examples of these structures. Combining this detector with tools to measure signaling near the cell surface, boundary motion, and other metrics, we measure how Kras and PIP2, two central signaling molecules, associate with blebs, a type of morphological motif. A Behind the Paper blog post details our motivation for building an automated analysis framework for 3D images.

Robust and automated detection of subcellular morphological motifs in 3D microscopy images

Blebs detected on an MV3 melanoma cell (representative of 19 cells), filopodia detected on an HBEC cell (representative of 13 cells), and lamellipodia detected on a dendritic cell (representative of 13 cells).