using a MRCA + snapper made mismatches with Taxon ~ TaxonSet labels

[drelo]

Dear BEAST users,

I have used BEAST2 before with other routines but I am stuck in this analysis.
I am using snapper to try to date a dataset using some secondary calibrations. I noticed the best approach to establish Species/Population in the Taxon Set tab was using the 'guess' option, so I relabeled the data to conform to this and I used the "everything before first '_'" . Since samples under the same species are labeled like this 'agu_H8' or 'agu_F5' etc, a Species label 'agu' is generated and it seemed to work, i.e. if I don't specify a taxon prior I can run this analysis just fine.

As soon as I include 2 sequences of agu + mk in a MRCA Taxon prior in the Priors panel the analysis fails.
For example I have these labels in the dataset

mk_B2
agu_A8
agu_H7

and the MRCA built has this

<distribution id="prior" spec="CompoundDistribution">
                <distribution id="AG.prior" spec="beast.base.evolution.tree.MRCAPrior" tree="@Tree.t:shorty">
                    <taxonset id="AG" spec="TaxonSet">
                        <taxon id="agu_A8_" spec="TaxonSet">
                            <taxon id="agu_A81" spec="Taxon"/>
                        </taxon>
                        <taxon id="agu_H7_" spec="TaxonSet">
                            <taxon id="agu_H71" spec="Taxon"/>
                        </taxon>
                        <taxon id="mk_B2_" spec="TaxonSet">
                            <taxon id="mk_B21" spec="Taxon"/>

and I got this error

validate and intialize error: Taxon <agu_A8_> could not be found in list of taxa. Choose one of [mob, olb, rz, oli, mk, agu]

If I use agu as a label for both I get another error

IDs should be unique. Duplicate id 'agu' found

Is there a way to fix this? I have no problem to manually edit the xml as a workaround now, but I haven't found a way to conciliate this.

Maybe this is related to this issue #957

Please find attached the xml generated moreprior.xml, the shorty.fasta file and the modified xml I tried to use replace.xml

shorty.txt
moreprior.txt
replace.txt

Thanks for your help.

Andrés

[rbouckaert]

Hi Andrés,

In snapper, there is only a single tree: the species tree, and the taxa of that tree are [mob, olb, rz, oli, mk, agu], so the MRCA prior should refer to these taxa, not the lineage taxa specified in the alignment. So, you were on the right way replacing agu_A8_ withagu. But since the agu taxon was already specified in the taxonset for the tree, an idref attribute should be used instead of an id attribute. If you replace in replace.txt

                    <taxonset id="AG" spec="TaxonSet">
                        <taxon id="agu" spec="TaxonSet">
                            <taxon id="agu_A81" spec="Taxon"/>
                        </taxon>
                        <taxon id="agu" spec="TaxonSet">
                            <taxon id="agu_H71" spec="Taxon"/>
                        </taxon>
                        <taxon id="mk" spec="TaxonSet">
                            <taxon id="mk_B21" spec="Taxon"/>
                        </taxon>
                    </taxonset>

with

                    <taxonset id="AG" spec="TaxonSet">
                        <taxon idref="agu"/>
                        <taxon idref="mk"/>
                    </taxonset>

it will run.

In BEAUti, there seems to be a problem with the MRCAPriorInputEditor picking up the correct taxon sets after they have been edited in the TaxonSetInputEditor in the snapper template, which is a BeastFX issues, so I'll create an issue there (CompEvol/BeastFX#50) and close this one.

[drelo]

Thank you very much! that fix made it work All the best, Andrés El mar, 18 abr 2023 a las 18:04, Remco Bouckaert ***@***.***>) escribió:

…

Hi Andrés, In snapper, there is only a single tree: the species tree, and the taxa of that tree are [mob, olb, rz, oli, mk, agu], so the MRCA prior should refer to these taxa, not the lineage taxa specified in the alignment. So, you were on the right way replacing agu_A8_ withagu. But since the agu taxon was already specified in the taxonset for the tree, an idref attribute should be used instead of an id attribute. If you replace in replace.txt <taxonset id="AG" spec="TaxonSet"> <taxon id="agu" spec="TaxonSet"> <taxon id="agu_A81" spec="Taxon"/> </taxon> <taxon id="agu" spec="TaxonSet"> <taxon id="agu_H71" spec="Taxon"/> </taxon> <taxon id="mk" spec="TaxonSet"> <taxon id="mk_B21" spec="Taxon"/> </taxon> </taxonset> with <taxonset id="AG" spec="TaxonSet"> <taxon idref="agu"/> <taxon idref="mk"/> </taxonset> it will run. In BEAUti, there seems to be a problem with the MRCAPriorInputEditor picking up the correct taxon sets after they have been edited in the TaxonSetInputEditor in the snapper template, which is a BeastFX <https://github.com/CompEvol/BeastFX> issues, so I'll create an issue there () and close this one. — Reply to this email directly, view it on GitHub <#1106 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACO2NCVPSL766GW76APHLE3XB36VFANCNFSM6AAAAAAXBPSZZQ> . You are receiving this because you authored the thread.Message ID: ***@***.***>