glennhickey
commented
2 weeks ago Sorry, I'm not very familiar with this type of data. If you want these regions in the alignment, I'd suggest leaving them and not masking them any differently than the rest of the assembly. If you don't want them in the alignment, then hardmasking may be simpler than cutting them out, especially if cutting them out will result in tons of tiny contigs.
Someone was asking recently about masking after aligning, which seems a bit related to what you may want to do as well, but I'm not aware of any tools that do this directly.
dylandebaun
Hello, I have some short-read genomes that were assembled via reference mapping (different references) that I would like to align to each other. I have consensus sequences with masked reference sites. We are discussing ways forward with this (since hard masking is inadvisable), either:
keeping these reference sites soft-masked along with soft-masked repetitive regions as input in cactus (since they hold some alignment information) OR
removing these regions from the fasta prior to cactus alignment (since they are not all repetitive regions and are likely incorrect as they were just 'carried over' from the reference [different species]).
Any insight on this? Is cactus treating soft masked as a repeat exclusively? Most papers I've seen only align denovo assemblies.