TPMS transfer detection (XD) : ouput improvments

[flassalle]

Hello,

TPMS XD works fine and (very very) fast for detecting systematically transfer events in a database of gene trees. But for exploitation of the results, and notably the mapping of the event in the input gene tree, one would need a more complete output that what is given presently.

For now, there are only the names of donor and receiver nodes in the species tree, but this is not straightforward to map it to the location of the event in the gene tree, and in many cases there are several solution because genes are present in multiple copies in genomes.

So, could you add the list of leaf labels in the transfered clade? that way ther is only one position where to map the event, that is the most recent common ancestor of leaves.

Thank you in advance!

Florent

[tbigot]

Hi Florant, and thanks again for your idea. It helps to improve the usability of this software and the scope of the results. I will investigate soon about the best way to do that and keep you up to date via this issue.

As a reminder, here is what you suggested me formerly, during an informal meeting:

familyName,index,donnorTaxon,(acceptorTaxonA,acceptorTaxonB)
;donnor: species1,species2,species3…
;acceptor: speciesA1,speciesA2
;acceptor: speciesB1,speciesB2

with speciesA1 and speciesA2 ∈ receiverTaxonA and respectively for other acceptor taxon (if suitable).

[flassalle]

I think it would be better having complete leaf labels instead of species names, because of the multiple matching problem I descibed, so the output sould be lookin more like this:

familyName,index,donnorTaxon,(acceptorTaxonA,acceptorTaxonB)
;donnor: species1,species2,species3…
;acceptor: speciesA1.prot1334,speciesA2.prot5975
;acceptor: speciesB1.prot1004,speciesB2.prot772

[tbigot]

You’re right, the solution I’ll implement must provide leaves’ names.

[tbigot]

Solved by commit 435ba75. Please try and re-open if it does not work as expected.