A deep learning framework for secondary structure annotation in cryo-EM maps
Copyright (C) 2020 Jiahua He
Software requirements:
Python (https://www.python.org) (ver. 3.7 or later)
Python package requirements:
pytorch (https://pytorch.org) (ver. 1.2 or later)
mrcfile (https://github.com/ccpem/mrcfile)
numpy (https://www.numpy.org)
tqdm (https://github.com/tqdm/tqdm)
The interpolation program "interp3d.f90" should be built as a python package 'interp3d' through f2py.
f2py -c interp3d.f90 -m interp3d
This command will generate an ELF file with name like "interp3d.cpython-*.so". Please keep "interp3d.cpython-*.so" with all python scripts "*.py" in the same directory.
In order to run python scripts properly, users should set the python path using one of the following ways.
-
adding python path to the header of each python script like this:
"#!/path/to/your/python" -
running the scripts with the full python path like this:
/path/to/your/python ./emnuss.py ...
Required files:
"emd_*.map": EM density map in MRC2014 format (download from EMDB).
"config.json": config file in JSON format. See "./6MRC/config.json".
The compressed trained models are stored in directory "./save/". Unzip them before use.
High resolution Example: EMD-9195
cd 6MRC/
../emnuss.py -mi emd_9195.map -mo sspred.mrc -t 0.45 --output
Middle resolution Example: EMD-3329
cd 5FVM/
../emnuss.py -mi emd_3329.map -mo sspred.mrc -t 0.044 --output
Attentions: a threshold value should be provided after flag "-t". For better visualization, please use the author recommended contour level (or half of it).
Output files: "sspred.mrc": secondary structure prediction of the given map (in MRC2014 format). "helix.mrc" and "sheet.mrc" and "coil.mrc": predictions of 3 secondary structure classes in individual MRC2014 files (use flag "--output").