Forum 02/11/2010

“Segger: A new watershed segmentation and fitting tool”

Thomas Goddard – Computer Graphics Laboratory, University California San Francisco

Segger is a segmentation and fitting tool for placing molecules in single-particle EM maps developed by Greg Pintilie and MIT.  It uses watershed segmentation combined with multiple levels of Gaussian filtering to produce a hierarchical segmentation.  Segger is a plug-in to UCSF Chimera and provides interactive control over the level of granularity of the segmentation.  It includes fitting capabilities that can align molecule principal axes to segmented density principal axes followed by local optimization and can search for combinations of segmented domains that match a molecule shape.  I’ll demonstrate Segger use on a bacteriophage lambda EM map.