Validating maps from single particle electron cryomicroscopy

• Published in: Current opinion in structural biology Vol. 34; pp. 135 - 144
• Main Authors: Rosenthal, Peter B, Rubinstein, John L
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.10.2015
• Subjects: Cryoelectron Microscopy - methods; Cryoelectron Microscopy - standards; Crystallography, X-Ray; Models, Molecular; Reproducibility of Results
• ISSN: 0959-440X; 1879-033X
• DOI: 10.1016/j.sbi.2015.07.002

•Single particle cryo-EM maps should be validated by specific tests.•Tilt-pair analysis shows that a map is consistent with image data at low resolution.•Map resolution should be consistent with map features and tests for over-fitting.•Maps at low or high resolution may be validated by fitting an atomic model.•Refinement of atomic models requires cross-validation tests to prevent over-fitting. Progress in single particle cryo-EM, most recently due to the introduction of direct detector devices, has made the high-resolution structure determination of biological assemblies smaller than 500kDa more routine, but has also increased attention on the need for tools to demonstrate the validity of single particle maps. Although map validation is a continuing subject of research, some consensus has been reached on procedures that reduce model bias and over-fitting during map refinement as well as specific tests that demonstrate map validity. Tilt-pair analysis may be used as a method for demonstrating the consistency at low resolution of a map with image data. For higher-resolution maps, new procedures for more robust resolution assessment and for validating the refinement of atomic coordinate models into single particle maps have been developed.