Support for GridION runs

[dfornika]

We will be increasing our usage of the GridION sequencer in the near-term.

To consider:
There is not (to my knowledge) a standard SampleSheet format for GridION runs that allows us to easily match up which sample used which barcode, and which sample belongs to which project.

It could be done with a simple .csv file with the following fields:

sample_id
barcode
project_id

...but that file would need to be deposited with a standard filename and in a standard location for each run.

The GridION also outputs a set of raw fastq files for each barcode, but it does not aggregate them into a single fastq file per barcode. That needs to be done by some downstream processing step. Should that happen before or after the run is uploaded? What should the directory structure and naming conventions be for those aggregated fastq files?