Wildcards in config.yaml for exclusion of contigs

[DaniEstevezBarcia]

This is a question regarding the section for the configuration of the reference genome. One solution I get to work with this is just by making a previous subset in the reference with samtools, but otherwise, I was wondering if there is any chance to use wildcards in the string lists in exclude, mito or sex-linked options.

I ask this because there are more than 1000 unclassified scaffolds on the reference genome I am working with.

Cheers,

[zjnolen]

Hey there!

This is a good idea, I've mostly worked with references where the minimum size threshold could be used to get rid of the unclassified ones, but there are likely many situations where that won't work without taking others away too.

Would an option to supply a file to each of these config lines be useful as a solution? I'm thinking just a text file that lists the scaffolds to drop for each category. This file could be generated with bash commands using wildcards and then be verified before going into the pipeline so that nothing accidentally gets included. My concern with a wildcard option directly in the config is that it would be harder for the user to know if something goes awry and extra scaffolds get included in the exclude list, since that happens internally. If a solution for this is still interesting for you, let me know if you have some thoughts on this!

Best,
Zach

[DaniEstevezBarcia]

Ah yes, you are right there. I was not sure you could provide a file. At the same time, you are also right about the size of the scaffolds, the min_size filter works (which also conveniently removes the mitochondrion... on this note, does it matter to provide the contig name for the mitochondrion if the min_size is set (and the mt is below that limit)?

Cheers,

[zjnolen]

The file option isn't currently there, but one I had in mind to add at some point for more fragmented assemblies if needed. It's something I plan to add, so I'll reopen this just to keep track of that, but glad the min_size option gets it working for your case!

Right now, there's no difference between removing the mitochondrion with the min_size option or the mitochondrion contig name, it's mainly there for those that don't want to put a min_size. It and the sex-linked list are also there as a potential route for future analyses to be possible on these, such as calling mitochondrial haplotypes or doing sliding window analyses on sex chromosomes without influencing genome-wide estimates or things like the PCA.