BUG: Incorrect SRA files (RNA-seq)

[simonvh]

Describe the bug

Some SRA files are incorrect (either due to submitter not supplying correct files, or SRA not having correct annotation). For example, this file:
SRR18972600 (GSM6077643). This marked as PE, but fastq-dump only gets one fastq file.

To Reproduce

RNA-seq workflow, with the following samples.tsv:

sample  assembly
GSM6077643     GRCh38

Expected behavior
I don't know! Either an informative error message (i.e. two fastq files should be there, there's only one, this can be checked) or the file being treated as single-end. The second option may not be correct, and I assume this may also be hard to implement. I would guess an error message would already be good, because the error is not easily identifiable from the log, but only when you manually repeat the steps.

One thing I noticed that these are not sequenced with Illumina, but with DNBSEQ-G400.

Low priority bug I would say, but something to be aware of.

[Maarten-vd-Sande]

Related to #228

We can't really fix this (we don't want) as it's not really our problem 😇

However what we can do is make the error message clear, which I think is actually a good idea 😄.

If DNBSEQ-G400 never works you can add it here so that seq2science reports a nice message.