Difference between revisions of "Alignment project"
Line 9: | Line 9: | ||
An alignment project can contain one or several references. In this case, we are running a [[multireference alignment]] project that operates simultaneously alignment and classification. | An alignment project can contain one or several references. In this case, we are running a [[multireference alignment]] project that operates simultaneously alignment and classification. | ||
− | |||
== Elements of an alignment project == | == Elements of an alignment project == | ||
Line 25: | Line 24: | ||
====Template==== | ====Template==== | ||
The initial density (or set of densities in a multiresolution project.), to be rotated multiple times against each one of the particles in the data folder. | The initial density (or set of densities in a multiresolution project.), to be rotated multiple times against each one of the particles in the data folder. | ||
− | |||
====Table==== | ====Table==== | ||
Line 35: | Line 33: | ||
=== Numerical Parameters === | === Numerical Parameters === | ||
− | They define the mathematical part of the project, in terms of sets of angles to scan, shifts to allow, used bandpass or binning options.... | + | They define the mathematical part of the project, in terms of sets of angles to scan, shifts to allow, used bandpass, [[multilevel refinement | angular refinement policy]] or binning options.... |
− | |||
=== Environment Parameters=== | === Environment Parameters=== |
Latest revision as of 08:41, 27 October 2017
The alignment project is one of the most important objects in Dynamo
An alignment project puts together all the elements needed to complete the alignment of a set of particles: the data particles, some auxiliary files (as template, masks, inititial orientations).
The project is also used to design the operations carried on the particles, to select and configure the environment where the actual computations will occur, to start the computations and to access and visualize the results.
An alignment project can contain one or several references. In this case, we are running a multireference alignment project that operates simultaneously alignment and classification.
Contents
Elements of an alignment project
Data
A set of files representing the particles cropped out of the tomograms. The easiest format to store particle data is the data folder.
Auxiliary Files
The auxiliary files needed for the computation can be designed by the users, or automatically created by Dynamo, depending on the level of apriori information that the user has and wants to set into the project.
Volumetric maps are accepted in .em, .spi or .mrc formats.
Template
The initial density (or set of densities in a multiresolution project.), to be rotated multiple times against each one of the particles in the data folder.
Table
The table expresses the metadata on the particles. A table is basically a matrix representing a particle at each row, each column representing a property (i.e., shifts, euler angles) according to a predefined convention
Masks
Dynamo uses several masks. The most important is the alignment mask. The classification mask is only relevant for multireference analysis. The option for defining a fourier mask on template is used in very few special cases. The smoothing mask is the one used for computing Fourier Shell Correlations.
Numerical Parameters
They define the mathematical part of the project, in terms of sets of angles to scan, shifts to allow, used bandpass, angular refinement policy or binning options....
Environment Parameters
The environment parameters are used to define where the project will be run, both in terms of hardware (CPU,GPU,multiple GPU, cluster of CPUs, how many cores) and software (using Matlab or using a standalone version of Dynamo)
Creation of an alignment project
Alignment projects are most easily created with the dcp GUI, that guides the user step by step. However, they can also be created from the command line using the command dvpr or dcp.new . You can get information on their syntax through
>> help dvpr
>> help dcp.new
Additionally, several utilities in Dynamo give you the option of creating projects from GUI interfaces.
Running an alignment project
Projects need to be checked and unfolded before they can be run. Once they are checked by Dynamo, a simple project script (execution script ) is produced, that can just be invoked in the command line or submitted to a queuing system. As the iterative process advances, many different results are created and stored in the project folder.
Running from the dcp GUI
You can launch the project in the dcp GUI pressing the run button. The GUI and the full Matlab desktop will be irresponsive during the execution of the projects (the message area of the GUI will be black). In the standalone, it is advisable to run the project in a different shell.
Running from the command line
In Matlab
Write the name of the project in the command line:
myProject
In the standalone
Working with the standalone, it is advisable to run the project in a shell where Dynamo has been activated (i.e., you have sourced the activation script) but not running (i.e., you are not under the Dynamo >> prompt). In such a system shell, you write the name of the execution script created during the unfolding:
./myProject.exe
in linux/Mac, or
myProject.bat
in Windows.
A step-by-step protocol for testing purposes can be found here.
Output of an alignment project
The main result is the average computed at the last iteration. Besides, Dynamo stores in the project folder the results (alignment parameters, averages and other intermediate results, as fourier occupancy maps)
Showing results of an alignment project
The most straightforward way of accessing results is to view the final average through the 'show' area of the dcp GUI.
In the long run, we find more comfortable to viewing results through the use of the ddb command. This utility extracts required items from the project tree, and it's specially useful when one wants to check not just the final average but also other intermediate and complementary results.