I'm trying to migrate the

fasta_index_dna

subworkflow to nf-test. This is proving trickier that I'd expect. @Matthias De Smet, you originally implemented this module by the looks of it. This subworkflow takes the fasta, but also an extra altliftover channel:

ch_fasta        // channel: [mandatory] [ val(meta), path(fasta) ]
        ch_altliftover  // channel: [mandatory, if aligner is bwamem or bwamem2 or snap] [ val(meta), path(altliftover) ]

The issue is in this section of the code:

BWAMEM1_INDEX.out.index
                .join(ch_altliftover)
                .map { meta, index, alt -> [meta, index + alt] }

which I believe is trying to add the altliftover file inside the folder that is output by bwamem. But I don't think we can use

+

to do this, because it is a file not a meta map, and what we get from nf-test begins (condensing and truncating for space):

"0": [[{"id": "test"},
                        ["/workspace/modules/.nf-test/tests/fcebacc24247d7e952390921d1be834a/workspace",
                         "/workspace/modules/.nf-test/tests/fcebacc24247d7e952390921d1be834a/modules",
                         "/workspace/modules/.nf-test/tests/fcebacc24247d7e952390921d1be834a/.nf-test",
                         "/workspace/modules/.nf-test/tests/fcebacc24247d7e952390921d1be834a/tests",
                         "/workspace/modules/.nf-test/tests/fcebacc24247d7e952390921d1be834a/fcebacc24247d7e952390921d1be834a",
                            [[[
                               ".command.begin:md5,d41d8cd98f00b204e9800998ecf8427e",

i.e. it returns a list of all the files and folders with full paths. If I actually give a file to ch_altliftover, it still gives full paths but only the new file not the original bwa folder at all. I don't know how to pass this file inside the folder, without adding an extra process.

Sorry, can you clarify a bit more what the intended output is from given inputs? I'm not quite understanding. If you're trying to add a file to folder, then keep in mind that if the folder is passed as a input to process, adding files to it will break reentrancy.

This is the current code Mahesh: https://github.com/nf-core/modules/blob/master/subworkflows/nf-core/fasta_index_dna/main.nf

Hi Simon, The point is to add the altliftover file to the output of the

*_INDEX

processes (in the case of

BWAMEM

and

BWAMEM2

) and to add it as input to the index process for

SNAPALIGNER

. All three of those support lifting alignments for alt contigs to the canonical chromosome, if the altliftover file is available

Yes, but I don't think the current code is functional as is

Can the liftover file not be an extra channel?

I think it has to sit in the same folder as the index, no?

I was just going to say that

shared a GIF

Basically something needs to change here I think. One of: • Add an extra process in the subworkflow to add the alt file into the bwa folder • Add an extra input to bwa index, to take the alt file and just pass it through into the bwa output folder • Add an extra input to bwa mem to use the alt file as a separate channel or element of the index channel

from the help of `bwamem2`:

-j            treat ALT contigs as part of the primary assembly (i.e. ignore <idxbase>.alt file)

Which led me to believe the the

*.alt

file needs to be in the same directory and use the same naming scheme as the rest of the index

Reminder of the bismark fasta issue: https://nfcore.slack.com/archives/C04PPJLT1T3/p1732288386864799?thread_ts=1732032677.595229&cid=C04PPJLT1T3

As far as I know, we have the alt in the bwa folders for GATK/GRCh38

So, what do we do with this subworkflow?

ok, so for me it's more a question of where do we want the alt to sit

It needs to go into the bwa/ folder

so for me it's more: • alt with build indexes • alt separated from build indexes Once we figure how to properly deal with this, we can adapt the subworkflow

igenomes has it in the index folder for GATK/GRCh38

I know, I put it there

You don't

which I don't like

Could it be made inside the bwa/index process?

we can always figure a way to make it inside

yeah, we could do either way, we just need to make a decision

that sounds perfect for a discussion on the maintainers meeting on Friday

Yeah, I'm at a board game convention