Phyloseq file is not generated about dadaist2

[telatin]

Hello and thanks for reportin @Jgruetzke

You might be right: if you check the log at the beginning you should have got a warning about sample names starting by a digit. It was only a warning because on some cases this is not an issue, but I want to fix this eventually.

Would you be please able to rerun it adding, for example, an "S" in front of the files? Thanks a million

[Jgruetzke]

Hi @telatin,

thank you very much for your fast response. I checked with the letter in front of the files, but still got the same error. However, it seems that the dashes in the sample names are a problem here. When I remove those, the phyloseq-file is generated. And also underscores should only be used for the _R1 and _R2, but not in other parts of the sample name, otherwise it is a problem for the metadata sheet generation.

The automatic sample naming from the fastq files in the metadatasheet using dadaist2-metadata is super convenient, but it would be esp. in this case quite useful to be able to use different sample names without changing the name of the fastq-files. I tested this by replacing the names in the #SampleID column of the metadata file and leaving the fastq names as it was, but it failed at the beginning of the pipeline.

[telatin]

Thanks for the feedback!
Yes, metadata is opinionated and the tradeoff between a reasonable metadata generated by the user (to be checked) and one automatically generated can be tricky. THere are flags to modify how the sample names is inferred (separator etc) but the idea would be that SampleName1_R1.fastq.gz would be the cleanest input for dadaist2.

A major rewrite is ongoing, if it sees the light metadata will be very different

[Jgruetzke]

I cross my fingers for the new version. I find it very useful!

[telatin]

I'll close it for now, but there will be some updates later this year about amplicon seq