This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Find structural variants comparing two assemblies

Is there a rapid way of comparing two large assemblies quantitatively, like generating a file with all large structural variants, or SV - gaps, inversions, etc., and theirs coordinates? The idea is basically to find out what one assembler particularly failed to assemble, compared to another one. Not using reads but assemblies alone, e.g. by breaking one of them on small pieces and use them as pseudo-reads.

There are ways of comparing assemblies, but they are not precise (e.g. dotplot, just visualizes but doesn't provide number and type of SV) and are finite (e.g. GMASS score or likehood test) - one cannot analyze what is different, but just HOW different.

This is simple task and quite typical for evaluation of assemblers' output, but I couldn't find any working solution. GATK finds only small SVs and poor at finding inversions, mummer's show-diff has the same problems + running two long on large genomes.

assembly genome comparative genomics

1 answer

Try these tools:

Assemblytics: https://github.com/MariaNattestad/Assemblytics

SVIM-asm: https://github.com/eldariont/svim-asm

SVMU: https://github.com/mahulchak/svmu

smartie-sv: https://github.com/zeeev/smartie-sv

Hope this helps!

Log in to answer this question.