Account for libraries that have alternative experiment data in merging workflow

[allyhawkins]

The current plan for merging objects is to make sure that any ADT data is still present in the final merged object. However, with multiplexed data, we do not want to include the HTOs or feature data. This is because we already have sample calls that we have added to the colData and having multiple clashing HTOs could be confusing. We are planning on implementing a flag in the function for merging SCE's to tell the function to return a merged object with merged altExps or without (see AlexsLemonade/scpcaTools#150 (comment)).

In order to pass that information along to the function mentioned (currently run in merge_sces.R), we will need to do the following:

• Grab the technology information for each library when reading in the library metadata
• Create a has_citeseq and is_multiplexed flag that can be passed to the merge_sce process.
• Add a flag for keep / don't keep altExp data to merge_sces.R and pass along to merge_sce_list.

I think in the actual workflow, we really only need to remove the altExp if it's multiplexed.
So then we should be able to create something like --remove_altExp flag that is used in merge_sces.R.
We could do the opposite of that and use has_citeseq ? --keep_altExp : ''. But I don't actually think we will need both.

merge_sces.R \
      --input_library_ids "${input_library_ids}" \
      --input_sce_files "${input_sces}" \
      --output_sce_file "${merged_sce_file}" \
      --n_hvg ${params.num_hvg} \
     ${is_multiplexed ? --remove_altExp : '' } \
      --threads ${task.cpus}

[allyhawkins]

Closed by #610