RELION ====== RELION is an open source cryoEM data processing suite. It is implemented here through SBGrid. Starting up ----------- There are two ways to run RELION on the Blanca cluster. - :ref:`sbatch submission` **Preferred** - :ref:`interactively` .. _sbatch submission: Sbatch submission ~~~~~~~~~~~~~~~~~ RELION is built with HPC integration in mind, so it can submit sbatch scripts for you and makes them easy to write inside the GUI. (underlying templates are found at ``/projects/biokem/software/biokem/users/src``). To start, allocate an interactive job with minimal resources to run the GUI on Blanca, start the GUI, and then ``Submit to queue``: .. code-block:: bash biokem-interactive cd /path/to/data/folder sbgrid relion Now in the ``Running`` tab: - Set ``Submit to queue`` to ``Yes`` - Defaults should be auto populated (``Walltime`` can be up to ``168:00:00``) - If you want to submit to Blana at large: - ``Queue name`` = ``blanca`` - ``QoS`` = ``preemtable`` - ``Account`` = ``blanca-biokem`` - ``Walltime`` = ``24:00:00`` (or less) You can submit multiple jobs at the same time using this method. You can use the GUI to moniter the output of each job. .. _interactively: Interactively ~~~~~~~~~~~~~ When starting interactively, you must allocated all of the resources you need before starting the GUI. Example: .. code-block:: bash biokem-interactive -c 24 -g 1 -m 48 cd /path/to/data/folder sbgrid relion The drawbacks with this method are: - Need to know how many resources you will use beforehand - Only way to allocate more resources is to exit screen and start over - Need to keep terminal alive while processing (24hr limit) - Have to wait until resources become available to start jobs - Sequesters idle resources from other users .. _RELION OTF: RELION OTF ~~~~~~~~~~ Similar to CryoSPARC's Live function, RELION can also preprocess data as it is being transferred into your folder using `Schedules` (See `here `_).