5.4 γ-TEMPy
γ-TEMPy uses a genetic algorithm to assemble component models
simultaneously in a cryo-EM map. The fitness function combines
the mutual information score (MI) (see Subheading 2.3) to quantify
the goodness-of-fit with a penalty score that helps to avoid clashes
between components. The script for running γ-TEMPy, which can
be run from the command line, can be found in the example scripts
from the TEMPy download. The necessary input to run γ-TEMPy
are: (a) coordinate file in PDB or mmCIF format. The file should
contain all the components (with chain IDs) separated with the
TER keyword. (b) Input density map in MRC file format, and
(c) the resolution of the map (A ˚ ). Optional inputs include the
number of solutions to generate, the number of generations for
the genetic algorithm to use, and the population size (number of
assembly fits). Since GA is a heuristic technique, multiple runs may
or may not produce similar results. The resulting output folder will
include the initial VQ points in a pdb file, a log file containing
details regarding the models generated, and the best scores
achieved, with the best model being output as separate pdb file.
5.5 Flexible Fitting
with Flex-EM
Flex-EM is available for download as a collection of python scripts
to be run from the command line (see Note 9) together with the
MODELLER software package [2]. The input parameters need to
be defined within the script “flex-em.py” to run Flex-EM. The
following parameters are essential: (a) The optimization method
must be set to “MD” (for simulated annealing molecular dynamics)
or “CG” (conjugate gradient minimization), (b) the input PDB
coordinate file that has been initially fit into the map (see Note 11);
(c) the cryo-EM density map file in “.mrc” or “.xplor” format;
(d) the voxel size (see Note 10), (e) the average resolution of the
input map, (f) the X, Y, Z origin of the map (see Note 12), the
number of CG iterations or MD runs, (g) The cap shift (maximal
atom movement allowed) must also be set (see Note 10). Additional input regarding the rigid bodies must also be supplied (see
Note 13). The output of Flex-EM contains PDB files for models
produced at each iteration of MD or CG, along with a final model
from the run, that can be used as input for another flex-EM run.
5.6 Density
Difference Mapping
A density difference mapping workflow is included with TEMPy
(“difference_map.py”) that can be run from the command line. The
input takes: (a) either two maps (for map-map difference) or a map
and atomic model (for map-model differences) (see Note 14),
(b) The resolution of both maps (in the case of map-map) or the
resolution of the map (in the case of map-model). The output is
given as two maps (map 1–map 2 and vice versa). Maps can be
visualized in Chimera and used as input for TEMPy.
5.7 Fitting Small
Molecules
For fitting small molecules, it is first necessary to obtain a collection
of physically plausible ligand conformations. This can be achieved
using ligand docking software such as GOLD [60] or AutoDock
CryoEM Density Fitting and Validation
215
γ-TEMPy uses a genetic algorithm to assemble component models
simultaneously in a cryo-EM map. The fitness function combines
the mutual information score (MI) (see Subheading 2.3) to quantify
the goodness-of-fit with a penalty score that helps to avoid clashes
between components. The script for running γ-TEMPy, which can
be run from the command line, can be found in the example scripts
from the TEMPy download. The necessary input to run γ-TEMPy
are: (a) coordinate file in PDB or mmCIF format. The file should
contain all the components (with chain IDs) separated with the
TER keyword. (b) Input density map in MRC file format, and
(c) the resolution of the map (A ˚ ). Optional inputs include the
number of solutions to generate, the number of generations for
the genetic algorithm to use, and the population size (number of
assembly fits). Since GA is a heuristic technique, multiple runs may
or may not produce similar results. The resulting output folder will
include the initial VQ points in a pdb file, a log file containing
details regarding the models generated, and the best scores
achieved, with the best model being output as separate pdb file.
5.5 Flexible Fitting
with Flex-EM
Flex-EM is available for download as a collection of python scripts
to be run from the command line (see Note 9) together with the
MODELLER software package [2]. The input parameters need to
be defined within the script “flex-em.py” to run Flex-EM. The
following parameters are essential: (a) The optimization method
must be set to “MD” (for simulated annealing molecular dynamics)
or “CG” (conjugate gradient minimization), (b) the input PDB
coordinate file that has been initially fit into the map (see Note 11);
(c) the cryo-EM density map file in “.mrc” or “.xplor” format;
(d) the voxel size (see Note 10), (e) the average resolution of the
input map, (f) the X, Y, Z origin of the map (see Note 12), the
number of CG iterations or MD runs, (g) The cap shift (maximal
atom movement allowed) must also be set (see Note 10). Additional input regarding the rigid bodies must also be supplied (see
Note 13). The output of Flex-EM contains PDB files for models
produced at each iteration of MD or CG, along with a final model
from the run, that can be used as input for another flex-EM run.
5.6 Density
Difference Mapping
A density difference mapping workflow is included with TEMPy
(“difference_map.py”) that can be run from the command line. The
input takes: (a) either two maps (for map-map difference) or a map
and atomic model (for map-model differences) (see Note 14),
(b) The resolution of both maps (in the case of map-map) or the
resolution of the map (in the case of map-model). The output is
given as two maps (map 1–map 2 and vice versa). Maps can be
visualized in Chimera and used as input for TEMPy.
5.7 Fitting Small
Molecules
For fitting small molecules, it is first necessary to obtain a collection
of physically plausible ligand conformations. This can be achieved
using ligand docking software such as GOLD [60] or AutoDock
CryoEM Density Fitting and Validation
215