make config parameters consistent across all steps

[erflynn]

Currently pre_qc uses different parameters, we want to make these all consistent so folks can use one config for multiple steps. Highest priority to make pre_qc and post_qc consistent; pre_fmx_qc and post_fmx_qc can come later.

[erflynn]

As far as I can tell the only different between original and v2 is:

original

"pools" : {
    "DM1" : {
      "libraries": {
        "TEST-POOL-DM1-SCG1": {
          "data_types": ""
        },
        "TEST-POOL-DM1-SCG2": {
          "ncells_loaded": 200,
          "data_types": ""
        }
      }

v2:

"pools": [
    {
      "name": "DM1",
      "libraries": [
        {
          "dir": "TEST-POOL-DM1-SCG1",
          "data_types": ["GEX"]
        },
        {
          "dir": "TEST-POOL-DM1-SCG2",
          "ncells_loaded": 200,
          "data_types": ["GEX"]
        }
      ]

is this correct?

[erflynn]

Tagging @dtm2451 so you have context for this @amadeovezz -- I do not think the libraries should be called dir -- that is the library name. Otherwise I am fine with the updates, we should just make them consistent for post_qc

[amadeovezz]

I do not think the libraries should be called dir

I originally used the word dir since this value is injected in path values - as the directory where the fastq files live.

My thought was, "as a user I should be able to point to a directory where my fastqs live"

I now see that library name = dir, which was confusing to me at first (and I think still is for new comers)

For this iteration, I think it is okay to change this key to name, given that this is a convention we use in the lab. In the future - I'd like to make the input file even more intuitive, and I think it is worth revisiting.

[erflynn]

good point, very open to updating the config to make it more intuitive.