DEPRECATED: How to run MICe build model on Niagara

For more information regarding MICe-build-model visit: https://wiki.mouseimaging.ca/display/MICePub/MiceBuildModelAlgorithm and https://github.com/CoBrALab/documentation/wiki/pydpiper:-Running-MBM.py-on-cross-sectional-data.

To long into a development node, run the following command (you will be prompted to enter your password:

ssh username@niagara.computecanada.ca

First, you will need upload your input files from you cic workstation into an input folder that you create on niagara.

To do so, you might run something like the following in a separate terminal:

rsync -avz *input_files.mnc username@niagara.computecanada.ca:/path/to/input_folder/

Lastly, you will need to upload any atlases you may want to use as your target atlas, or for a subsequent MAGeT analysis up on niagara (in the same way as above).

Once your files have been uploaded, you will need to create a directory in which to run your MICe-build-model analysis on niagara, for example: mbm_run_date.

Next, you will need to create a run script, for example MBM_run_script.sh, which can be written using nano. Here is an example:

module load cobralab

MBM.py --verbose --pipeline-name=whatever_name_you_want \
--files path_to/input_folder/*input_files.mnc \
--resolution resolution_of_your_data \

MBM.py: the pipeline that will run the registrations. Running MBM.py --help in your terminal will give you a list of all the pipeline options.

--verbose: will print jobs to screen

--pipeline-name: whatever name makes sense to you

--resolution: voxel resolution of input data

--stats-kernels: determines your blurring kernel, which should be double what your resolution is, so for 40um isotropic data, it would be 0.08, for a 100um isotropic data it would be 0.2

There are many more options you can play around with, this is a starting off point.

To actually run your script, it is recommended to use tmux. To do so, run tmux in your command line, following by bash MBM_run_script.sh

Once your run is complete, choose the files you will need to do statistics, and pull them down to your cic workstation. This might include your relative and absolute Jacobian determinant files, the nlin-3 average file and mask, as well as the lsq6 files, and MAGeT label files.