Resolving repetitive(?) region in animal mitochondrion

[brettmorg]

Hello!

I am assembling a beetle mitochondrial genome, using this mt genome from the same beetle family as the seed. It appears mostly assembled, except for one region that has many small repeats or fragments. I'm wondering if you have any suggestions to obtain a circular assembly. I've tried playing around a bit with -w and --reduce-reads-for-coverage, with no better results.

Log: get_org.log.txt

Thanks very much!

[Kinggerm]

Repeats in beetle mitogenomes are difficult. It's beyond the data capability of short read sequencing. AFAIK, increasing kmer helps resolving some short tangled nodes but can't get you a circular assembly. Replacing those uncertainties with Ns (gaps) is safe and reasonable, and what I will do for my study. Or, you can do long-read sequencing to solve those repeats.

Hope it helps, Jianjun

[brettmorg]

Great, I actually do have nanopore data from this specimen. Do you know of a way to use long reads to resolve the repeats directly from the GetOrganelle results? Or would I need to go for a hybrid approach with a different assembler?

Cheers, Brett

[Kinggerm]

That's cool.

I've conducted some simple tests on organelle genome assembly using flye/canu based on ONT/SMRT reads. The result is not promising even after polish, e.g. the IRs in chloroplast genome are not identical (should be). You can try other assembler if you find a suitable one.

I'm developing another tool using the accurate assembly graph (based on short read) and long reads to navigate the accurate paths and estimate the proportions of the paths. It will come soon. I can keep you posted if you are interested.

[brettmorg]

Yes, that sounds perfect! Please do let me know.

[Andreas-Bio]

I also subscribed to this thread to get to know the latest updates in this matter.