Harshil Patel

Thanks @ggabernet! Yes, it would be awesome to get a prototype together for this as soon as we can. The command @ewels suggested [here](https://github.com/nf-core/viralrecon/pull/174#issuecomment-830651227) sounds 👌🏽if we decide to add...

Related to https://github.com/nf-core/tools/issues/1155

As it stands `bioconda::idr` and `bioconda::macs2` wont build in the same environment because they are built with Python 3 and Python 2, respectively. https://anaconda.org/bioconda/idr/files https://anaconda.org/bioconda/macs2/files Need a workaround for this.

Hi @chbk ! Thank you for creating this issue. This is quite odd because the the full-tests we run on pipeline release have the correct sections in the [MultiQC report](https://nf-core-awsmegatests.s3-eu-west-1.amazonaws.com/atacseq/results-1b3a832db5a53c92c2c89a4d6d79455e860461ad/multiqc/broadPeak/multiqc_report.html#picard)....

Ok. Deleted the last comment because it was getting messy but I see the problem now...unless we default to `find` the first instance of the index in that structure? Is...

We could but how would having an extra fasta file being staged here help? We would still have the same issues with paths no? I have to say the biggest...

Yup, but that name would still be found 3 times in the example structure above?

Ok. Then I think we will still need some sort of hack in order to use tarred directories containing indices? That is where the `find` is the most useful because...

Yep, I did it without permission because it made sense at the time and it solved quite a few problems with passing the index around and standardising the use of...

Ok. So if you are able to get a working prototype together where we *don't* have to pass the fasta file to the alignment process and where we can use...