The routine mtriage will evaluate the resolution in a map, compare half-maps, and provide some basic statistics about the map
If only one map is provided, mtriage will calculte some basic map stitistics (min, max, mean, standard deviation), show histogram of map values and calculate resolution estimate d99.
If two half-maps are given in addition to the full map, then FSC curve between the two half-maps will be calculated as well as resolution estimate d_FSC(0.143).
If an atomic model is provided then two sets of statistics will be calculated: using original and masked maps. The soft mask is calculated using the atomic model as described in citation below. Also, some model-based resolution estimates will be calculated, such as d_model and d_FSC_model as well as map-model FSC curve.
Model file can be in PDB or mmCIF formats. Map file is expected to be CCP4, mrc or other related format.
Guide to resolution estimates:
- d_FSC - highest resolution at which the experimental data are confident. Obtained from FSC curve calculated using two half-maps and taken at FSC=0.143.
- d99 - resolution cutoff beyond which Fourier map coefficients are negligeably small. Calculated form the map.
- d_model - resolution cutoff at which the model map is the most similar to the target (experimental) map. Requires map and model. For d_model to be meaningful, model is expected to fit the map as good as possible.
- d_FSC_model - resolution cutoff up to which the model and map Fourier coefficients are similar.
Refer to Table 2 in article below for more details about interpretation of various resolution estimates.
Running mtriage is easy. From the command-line you can type:
phenix.mtriage map.mrc phenix.mtriage map.mrc half_map_1.map half_map_2.map phenix.mtriage map.mrc model.pdb phenix.mtriage map.mrc model.pdb half_map_1.map half_map_2.map
All inputs maps are expected to be on the same grid, the program will stop otherwise. Box information (aka unit cell CRYST1 in PDB files, for example) must match across all inputs.