-
Notifications
You must be signed in to change notification settings - Fork 4
DeepRes local resolution
DeepRes is a method to estimate a new local quality measure for 3D cryoEM maps that adopts the form of a local resolution-type of information. DeepRes is based on deep learning 3D features detection.
Two networks were previously trained to detect resolutions range between 2.5 and 13 A and between 1.5 and 6 A for very high resolution maps.
DeepRes is part of the Xmipp processing package and integrated in the image processing framework Scipion. Therefore, to use DeepRes the user must install Scipion and then install the Xmipp plugin. At the moment (until the new version of Xmipp comes out) the software is in the devel development branch. To access this development branch, first, the user must install Scipion from GitHub (see here). Afterwards the user must install the xmipp plugin, but in developer mode to have access to the new developments (see here).
Once installed, the user must download from scipion the deep learning tools (which contains all the necessary libraries and previously trained networks)
./scipion installb deepLearningToolkit
In the Scipion GUI, the protocol can be found in the Resolution branch, with the name: local deepRes.
Input and output
The algorithm requires as input a 3D Cryo-EM density map and a mask enclosing the macromolecule, and as a result a resolution map is obtained. Different visualization options have been implemented into Scipion to analyze the results (for example: the local resolution map, a resolution histogram, colored slices of the resolution map and the display of the original map in Chimeera colored according to the obtained resolution values).