A scripting language approach to control software for cryo-electron microscopy.

TitleA scripting language approach to control software for cryo-electron microscopy.
Publication TypeJournal Article
Year of Publication2002
AuthorsBaxter, B, Leith, AD, Frank, J
JournalProc IEEE Int. Symp. Biomed. Imag.
Start Page301

Cryo-electron microscopy (Cryo-EM) of single particles has developed into a widely used technique for determining the 3-dimensional structure of large molecules and molecular assemblies. The low signal-to-noise ratio of cryo-EM requires thousands of images of single molecules to be averaged together. The field has matured to the point where realization of high-resolution structures is limited primarily by computational constraints. These constraints are at the algorithmic level, as well as the control level, where dozens of complex procedures and thousands of mathematical operations are applied to the raw data. SPIDER is a highly modular and flexible software package for single particle reconstruction. A typical reconstruction project involves dozens of procedure files, which in turn draw on hundreds of available low-level operations. In the present system, it is virtually impossible to rerun the system from selected branching points in the process flow to determine the effects of various parameters values. A Reconstruction Engine (RE) is being developed as a high-level "shell" for controlling processes in the SPIDER software system. The RE allows backtracking, optimization of parameters, and automation of processing flow. The RE is implemented in a scripting language, Python, which provides an overall management capability at the global level of the project.


You are here