Assessing Map Quality (cryo-EM)

Why

The first step after data processing is analysis to assess data quality and detect any pathologies that might thwart structure solution. The quality of the reconstruction, and therefore the interpretability of a cryo-EM map, can deteriorate from many causes including structural heterogeneity, radiation damage, and beam-induced sample movement.

Data quality is typically expressed by the resolution, but this term has a different meaning for cryo-EM than crystallographic experiments. For a crystallographic data set, resolution depends on the largest angle to which diffracted beams were measured (or equivalently, the shortest distance between reciprocal-lattice planes). For a cryo-EM map, the overall resolution (d_FSC) is usually defined as the maximum spatial resolution at which the information content of the map is reliable.

The cryo-EM resolution is obtained by analyzing the Fourier shell correlation (FSC) for two cryo-EM half-maps binned in resolution shells. If the macromolecule is structurally heterogeneous, a single resolution value is adequate. 'Local resolutions' are also assigned to different map regions.

How

The phenix.mtriage program estimates the resolution of cryo-EM maps using several different approaches and then provides a summary of the map statistics. The tool requires only a full cryo-EM map (CCP4, mrc, or other related format). However, you can obtain additional information by inputting a file of two half-maps, an atomic model (PDB or mmCIF format), or an atomic model and two-half maps.

How to use the phenix.mtriage: Click here

Common issues

Program stops: All input maps are expected to be on the same grid. Also, box information (e.g., unit cell of CRYST1 record in PDB files) must match across all the inputs.

Related programs and Documentation

• phenix.show_map_info: This program lists the properties of a map, including the origin, grid points, unit cell, and map size.
• phenix.map_box: This program cuts out a box around selected atoms from a larger map (MTZ, CCP4, or X-plor format).
• phenix.map_symmetry: Some molecules (e.g., viruses) have high internal symmetry, so you can reduce the map to the repeating unit. This program finds such symmetries, and phenix.map_box can extract the unique part of the map (using the extract-unique option).
• phenix.combine_focused_maps: This program creates a weighted composite map from a set of locally focused maps and associated models, where each part of each map is weighted by its correlation with the corresponding model.