100s of chimeric mitochondrial genome contigs and same region duplicated in 2 contigs

[rb-commits]

[rb-commits]

Oops, forget to add the dot plots for the duplicated region. Here it is.

[lh3]

Hifiasm may produce different copies of mitochondria. That may be caused by heteroplasmy. However, in general, reference-based analysis may not be a good way to understand a new sample:

• If you want to show the two mitochondria are misassembled, you need to confirm with reads.

• Your sample is fairly diverged from the known reference. The "overlap" could just be segmental duplications. You need coverage plot and/or breaking reads to investigate.

• BUSCO duplicates can be misleading. There are papers showing this.

Also a question: are you sure the input yeast samples come from nearly identical strains? Hifiasm is so sensitive that it will put tiny differences between strains into different contigs.

[rb-commits]

Thanks for the quick response. This organism is dikaryotic (i.e. they have two separate nuclei), with a high level of heterozygosity between two haplotypes (0.84 SNP/Kb). I was wondering if those two contigs (in the left) are actually haplotypes (see the dot plots).

There are not only 2 contigs of mitochondria but in 100s. By the way, the longer (95kb) mitochondrial contig is also from the same Hifiasm assembly, not the reference. I haven't checked the read depth for those mitochondrial contigs. I will check it and will let you know.

[HenrivdGeest]

For plants I've seen even a thousand contigs being partly the chloroplast. I reasoned that because of excessive coverage of the plastids, even common or rare systematic errors end up in different contigs. ( among heterogeneity in the sample itself). The coverage according to hifiasm is mostly 0 or 1 on these contigs. Maybe you have similar issues with the mitochondria? What if you map all your input hifi data to a single contigs that represents the mtDNA?

…

On Wed, 27 Jan 2021, 22:47 rb-commits, ***@***.***> wrote: Thanks for the quick response. This organism is dikaryotic (i.e. they have two separate nuclei), with a high level of heterozygosity between two haplotypes (0.84 SNP/Kb). I was wondering if those two contigs (in the left) are actually haplotypes (see the dot plots). [image: image] <https://user-images.githubusercontent.com/78109099/106058147-78af4980-60b6-11eb-95b8-332a4b9b2845.png> There are not only 2 contigs of mitochondria but in 100s. I haven't checked the read depth for those mitochondrial contigs. I will check it and will let you know. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#68 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AARZCFIZJB753EX3TSJWAJDS4CCY3ANCNFSM4WV4OFKQ> .