addition, one may want to run the RELION “Local resolution” job-type to generate maps that are filtered according to
their local resolution.
19. After interpretation of each of the individual body reconstructions in terms of atomic models, one may want to generate a
composite atomic model of the entire complex. One way to do
this is to perform rigid-body fitting of the individually refined
atomic models for each body in the initial consensus reconstruction. However, it is crucial to understand that a single
atomic model can never be a good description for the flexible
complex. In particular, the assumption that each body moves as
a complete rigid body is most likely flawed at the interfaces
between the different bodies. This is because local conformational changes are likely to occur at the interface as one body
moves relative to its neighbor. To some extent, overlapping
body masks may help in the modeling of the interfaces, but
often there may simply be insufficient information in the data
set to define the multiple different conformations at the interfaces between bodies. Therefore, one should be particularly
careful with refinements of composite models in the consensus
map, which may sometimes seem necessary to handle clashes at
the interfaces.
20. Note 19 raises the question of which maps to submit to the
Electron Microscopy Data Bank (EMDB) and which models to
submit to the Protein Data Base (PDB) [13]. Multiple options
are possible, and the field has not yet converged on one. An
important consideration is that the maps used for building all
parts of the atomic model are preserved. For example, one
could make a separate EMDB and PDB entry for each body,
together with separate EMDB and PDB entries for the consensus map and the composite atomic model. Alternatively,
one could make one EMDB and one PDB entry with the
consensus map and the composite model, and then submit
reconstructions for the individual bodies that were used for
model building to the EMDB as supplementary maps. For each
post-processed map used for atomic modeling, it is strongly
recommended to submit the unfiltered half-maps and the
masks used for body-definition and/or post-processing as supplementary maps to the EMDB.
21. Movies can also be generated in Pymol, where the ten maps
should be loaded into different states of a single map object.
22. From RELION release-3.1 onwards, partial signal subtraction
after multi-body refinement is different from the one described
in
Subheading
3.3.
Instead
of
using
the
“relion_flex_analyse” program, the optimiser.star
file from the multi-body refinement can be given directly as
158
Takanori Nakane and Sjors H. W. Scheres
their local resolution.
19. After interpretation of each of the individual body reconstructions in terms of atomic models, one may want to generate a
composite atomic model of the entire complex. One way to do
this is to perform rigid-body fitting of the individually refined
atomic models for each body in the initial consensus reconstruction. However, it is crucial to understand that a single
atomic model can never be a good description for the flexible
complex. In particular, the assumption that each body moves as
a complete rigid body is most likely flawed at the interfaces
between the different bodies. This is because local conformational changes are likely to occur at the interface as one body
moves relative to its neighbor. To some extent, overlapping
body masks may help in the modeling of the interfaces, but
often there may simply be insufficient information in the data
set to define the multiple different conformations at the interfaces between bodies. Therefore, one should be particularly
careful with refinements of composite models in the consensus
map, which may sometimes seem necessary to handle clashes at
the interfaces.
20. Note 19 raises the question of which maps to submit to the
Electron Microscopy Data Bank (EMDB) and which models to
submit to the Protein Data Base (PDB) [13]. Multiple options
are possible, and the field has not yet converged on one. An
important consideration is that the maps used for building all
parts of the atomic model are preserved. For example, one
could make a separate EMDB and PDB entry for each body,
together with separate EMDB and PDB entries for the consensus map and the composite atomic model. Alternatively,
one could make one EMDB and one PDB entry with the
consensus map and the composite model, and then submit
reconstructions for the individual bodies that were used for
model building to the EMDB as supplementary maps. For each
post-processed map used for atomic modeling, it is strongly
recommended to submit the unfiltered half-maps and the
masks used for body-definition and/or post-processing as supplementary maps to the EMDB.
21. Movies can also be generated in Pymol, where the ten maps
should be loaded into different states of a single map object.
22. From RELION release-3.1 onwards, partial signal subtraction
after multi-body refinement is different from the one described
in
Subheading
3.3.
Instead
of
using
the
“relion_flex_analyse” program, the optimiser.star
file from the multi-body refinement can be given directly as
158
Takanori Nakane and Sjors H. W. Scheres