Manage DNA-RNA combinations in the consensus through sample name instead of patient

Description of feature

At the moment the combination between DNA-RNA samples at the consensus module of the pipeline is done through patient in the samplesheet but in more complex scenarios where there is more than one sample per patient this does not work. Ideally, DNA and RNA should be done through sample name to make sure we assign the proper pairs together.

To dos:

• Revise code for DNA-RNA combinations in the vcf_consensus subworkflow
• Change patient combinations for sample combinations
• Review ids of filenames and how outputs are named (many are named through sample as well and could cause conflicts in DNA and RNA have same name).
• Review samplesheet: potentially add data_type = RNA/DNA to samplessheet instead of identifying RNA by the status, i.e. status should only define if tumour or normal and data_type of DNA or RNA.

This comes from a very useful discussion with @tdanhorn , thank you.

Note:
Current combinations when you have several samples for same patient is unclear due to lack of testing and test data. A workaround is to treat the desired combinations as different patients. For example, patient1 might become patient1_A , patient1_B , patient1_C, etc. A very big flaw of this workaround is that you will need to specify the normal sample for each patient as well, making things redundant if you want to use the same normal for most samples.Alternatively, one could submit the pipeline separately for each patient set and hit -resume every run so the normal processes are cached and only the tumours will be done. Not ideal but hopefully this will get solved in time. If you have any feedback or what to propose any solutions please feel free to comment.