Mitochondrial DNA is longer than expected

[Phismil]

Dear Jianjun Jin
Thank you for maintaining the repository.
I am assembling the mitogenome from a wild cat species, and the size is expected to be ~16K.
However, my assemblies are constantly more than 24K. I tried different Kmers.
Is there any way to define an upper limit for the assembly?
I will appreciate any advise in advance

[JianjunJin]

• Please always provide a complete log file.
• For your question, a slimmed assembly graph (.extend.fastg in the output directory) is necessary for troubleshooting.

[Phismil]

Dear Jianjun Jin
Thank you for your kind response.
Please find attached the requested files.
Regards

extended_K105.assembly_graph.fastg.extend-animal_mt.fastg.txt
get_org.log.txt

[JianjunJin]

Hi, I deeply apologize for my slip. I must have checked it but forgot to respond.

This is an interesting sample exhibiting "heteroplasmy" in the assembly graph. The depths of the two variants are so even that none of them can be treated as contamination by GetOrganelle, which is why a 24k-chimeric result was generated.

It's difficult to explain why this happens. It may be NUMT or contamination. We observe similar "heteroplasmy" in the graph in our benchmark test (of big cats too!! see here). You probably know better than me. Looking forward to your opinion on your dataset.

Good luck!

[imogen-foote]

Hi, did this issue get solved? I am having a similar problem with get_organelle_from_assembly.py where my mitochondrial genome is coming out around 38Kb but I am expecting ~19K.
I'm sorry I don't seem to be able to attach the .gfa file format of the slimmed assembly graph, but if you're able to tell me how I can add that too.
get_org.log.txt

[imogen-foote]

Thanks @JianjunJin. I tried that (below code) but it's given exactly the same output as before.

get_organelle_from_assembly.py  -F animal_mt \
                                -g $read_dir/assembly_graph.gfa \
                                -o $output_dir \
                                -t 10 \
                                --overwrite \
                                --verbose \
                                --min-depth 48

I'm having trouble getting my annotation pipeline to work as well so can't check that just now. Is there anything else I could try?

[JianjunJin]

Could you visualize the big graph with Bandage and the csv file loaded?

[imogen-foote]

Yes, see below. Is there anything in particular that I should be looking for?

[JianjunJin]

Thanks. I see.
The 38Kb-single contig seems to include the real mt and non-target sequence too. So it is likely to be the disassembly of flye. Redoing the flye assembly may be an option. BTW, what type of data did you use?

[imogen-foote]

Aaah ok. I did the flye assembly with ONT data. I tried the assembly with different quality filtering thresholds (was supposed to be Q20+ data but on average came out a little lower than that) and all of them look odd for the mitogenome but the nuclear looks good. I also have some illumina data so I might try the get_organelle_from_reads.py with that and see if that produces a better mitogenome.
Thanks so much for you help with this issue and for the tool!