Add check for unique lanes in sample when parsing TSV files from FASTQ samples

[maxulysse]

Issue by @maxulysse, moved from SciLifeLab#626

We should have a warning when at least one sample has more than 2 FASTQ with the same lane.
Right now, it's failing at the merging BAMs step.

Describe the solution you'd like

Add some checks when parsing the TSV file

Describe alternatives you've considered

No alternatives considered at the moment

Additional context

Issue discovered by Katja when running some WES project

Issue by @nicweb, moved from SciLifeLab#714

When there is no lane information given in the sample.tsv markduplicates will fail with

htsjdk.samtools.SAMFormatException: Error parsing SAM header. Problem
parsing @rg key:value pair.

as readgroup ID is empty.

Would be nice to check TSV before execution or autoset empty lane with sampleID or similar to avoid downstream execution halt, or update documentation that field should not be empty.

[FriederikeHanssen]

I'll reopen it, this was not actually closed.

[asp8200]

Message from @maxulysse in Slack-thread Sarek-3.0 (https://nfcore.slack.com/archives/C02MDBZAYJK/p1656668331195689): for the uniqueness of lanes and samples.
For a single patient, we can have several times the same sample as long as they have different lanes
And I think a sample should be unique (we should not have the same sample in different patients)

[asp8200]

This GitHub-issue seem to consists of two parts (SciLifeLab#626 and SciLifeLab#714). The latter part may already have been fixed. I tried running dev-Sarek with an input csv-like this:

patient,sex,sample,lane,fastq_1,fastq_2
NA12878,XX,NA12878,,/faststorage/home/aspe/test_data/85gnb0878-NA12878-DNA_Blood-WGS_v2_SUB_R1.fastq.gz,/faststorage/home/aspe/test_data/85gnb0878-NA12878-DNA_Blood-WGS_v2_SUB_R2.fastq.gz

that is, one where the lane-field is just empty, and Sarek crashed with the error message :

WARN  nextflow.Nextflow - Missing or unknown field in csv file header. Please check your samplesheet

I got the same error-message when removing the lane-column from the csv-file:

patient,sex,sample,fastq_1,fastq_2
NA12878,XX,NA12878,/faststorage/home/aspe/test_data/85gnb0878-NA12878-DNA_Blood-WGS_v2_SUB_R1.fastq.gz,/faststorage/home/aspe/test_data/85gnb0878-NA12878-DNA_Blood-WGS_v2_SUB_R2.fastq.gz

Isn't this behaviour okay?