Run poseidon with mitochondrial sequences

[pedronachtigall]

Hi @hoelzer,

First, thank you for designing this nice tool.
I want to use poseidon in a set of mitochondrial genes, but, apparently, poseidon only uses the Standard Code.
Is there an option to decide the translation table used (e.g., set the Vertebrate Mitochondrial Code)?

Best regards,
Pedro

[hoelzer]

Hi @pedronachtigall

thx for your interest in the pipeline. Yes, at the moment PoSeiDon only uses the standard code. I think there are different steps where I would need to adjust that (codon-aware alignment via TranslatorX, CODEML, ...) but should be possible. I will set it on my todo but unfortunately, this will need some time.

[pedronachtigall]

Hi @hoelzer

Please, let me know once you have implemented this into poseidon.

Best,
Pedro

[etwatson]

Hi, I would also be interested in using poseidon for mitochondrial sequences. Have you implemented alternative genetic codes yet? I would be using the invertebrate mitochondrial code as well as the mold and yeast mitochondrial codes.

[hoelzer]

Hey, uff, a three-year-old issue 👀

Sorry, I do not have time to add new features to PoSeiDon.

But I quickly checked what would need to be adjusted. I think

Codon-aware alignment: https://github.com/rnajena/poseidon/blob/master/modules/translatorx.nf#L14

translatorx has a parameter:

  -c: genetic code to translate the sequences (Optional). Available options are:
      1   Standard
      2   Vertebrate Mitochondrial
      3   Yeast Mitochondrial
      4   Mold Mitochondrial; Protozoan Mitochondrial; Coelenterate Mitochondrial; Mycoplasma; Spiroplasma
      5   Invertebrate Mitochondrial
      6   Ciliate Nuclear; Dasycladacean Nuclear; Hexamita Nuclear
      9   Echinoderm Mitochondrial; Flatworm Mitochondrial
      10  Euplotid Nuclear
      11  Bacterial and Plant Plastid
      12  Alternative Yeast Nuclear
      13  Ascidian Mitochondrial
      14  Alternative Flatworm Mitochondrial
      15  Blepharisma Macronuclear
      16  Chlorophycean Mitochondrial
      21  Trematode Mitochondrial
      22  Scenedesmus obliquus Mitochondrial
      23  Thraustochytrium Mitochondrial
      100 Ancestral Arthropod Mitochondrial Code (AGG=K)
      101 Hemichordate Mitochondrial
      Default: 1 (Standard code)

Should be easy.

Recombination detection via GARD: https://github.com/rnajena/poseidon/blob/master/modules/gard.nf

Apparently GARD also has an option to switch to different genetic codes:

https://www.datamonkey.org/gard

But I can not see how to easily change the code.

**Probably more difficult. **

The actual detection of sites under positive selection with CODEML: https://github.com/rnajena/poseidon/blob/master/modules/codeml.nf

According to https://academic.oup.com/mbe/article/40/4/msad041/7140562 one can change the genetic code

whereas the genetic code is specified with the variable icode

Ah yeah, found it:

https://github.com/rnajena/poseidon/blob/master/bin/codeml_built.rb#L47

This post describes all available genetic code options for CODEML:

https://www.biostars.org/p/322110/

Should be medium effort to implement

---

I think these are the three constructions sides to implement that.

Maybe you are already fine with having 1) and 3) if recombination is not as important for you (although one should usually consider recombination events in selection analysis).

I can not promise, but will see if I find a time slot to add that.

[hoelzer]

Code number comparison between translatorX and CODEML

Code (TranslatorX)	Genetic Code	Code (CODEML)
1	Standard	0 (Universal)
2	Vertebrate Mitochondrial	1 (Mammalian Mitochondrial)
3	Yeast Mitochondrial	2 (yeast mt)
4	Mold Mitochondrial; Protozoan Mitochondrial; Coelenterate Mitochondrial; Mycoplasma; Spiroplasma	3 (mold mt)
5	Invertebrate Mitochondrial	4 (invertebrate mt)
6	Ciliate Nuclear; Dasycladacean Nuclear; Hexamita Nuclear	5 (Ciliate Nuclear)
9	Echinoderm Mitochondrial; Flatworm Mitochondrial	6 (Echinoderm Mitochondrial)
10	Euplotid Nuclear	7 (Euplotid Mitochondrial)
11	Bacterial and Plant Plastid	N/A
12	Alternative Yeast Nuclear	8 (alternative yeast nu)
13	Ascidian Mitochondrial	9 (ascidian mt)
14	Alternative Flatworm Mitochondrial	N/A
15	Blepharisma Macronuclear	10 (Blepharisma Nuclear)
16	Chlorophycean Mitochondrial	N/A
21	Trematode Mitochondrial	N/A
22	Scenedesmus obliquus Mitochondrial	N/A
23	Thraustochytrium Mitochondrial	N/A
100	Ancestral Arthropod Mitochondrial Code (AGG=K)	N/A
101	Hemichordate Mitochondrial	N/A

[hoelzer]

@pedronachtigall @etwatson I implemented basic functionality for genetic code switches:

https://github.com/rnajena/poseidon/releases/tag/1.1.0

See the new --help message.

I did not test this extensively but at least you can now change the genetic code for TranslatorX (codon-aware alignment) and CODEML (selection detection).

What is missing is GARD (recombination detection) which currently still runs based on the standard genetic code.

[etwatson]

Martin, I appreciate your attention for this older issue and for producing a working solution. I will give it a try and see if it works for me. I still need to run GARD, believe it or not we do see regular mitochondrial recombination in our organism, so I will need to work out a solution for this.

Thanks again!

[hoelzer]

@etwatson pleasure, hope it helps!

regarding GARD: ah okay, as I wrote above it is possible to change the genetic code also for GARD but that seemed way more complicated then what I just implemented for the alignment and selection steps. If you find a way or have any hints regarding how to modify GARD (and what the genetic codes are), I can see if I can add that relatively fast

[pedronachtigall]

Hi @hoelzer ,

Thank you for your attention with this issue. It is a great implementation and it will be very useful.
I am not working with mitochondrial sequences anymore, but I will certainly give it a try in further analyses.

Best,
Pedro