directional RRW in continuous phylogeography

44 views
Skip to first unread message

Ksenia Safina

unread,
Sep 19, 2022, 2:04:08 PM9/19/22
to beast-users
Hello,

I am trying to use BEAST1.10.4 to run a continuous phylogeography analysis assuming a directional RRW. I selected this option in Beauti, which I think produced this xml block:

<multivariateTraitLikelihood id="location.traitLikelihood" traitName="location" useTreeLength="true" scaleByTime="true" reportAsMultivariate="true" reciprocalRates="true" integrateInternalTraits="true">
        <multivariateDiffusionModel idref="location.diffusionModel"/>
        <treeModel idref="treeModel"/>
        <traitParameter>
            <parameter id="leaf.location"/>
        </traitParameter>
        <driftModels>
            <strictClockBranchRates>
                <parameter id="location.driftRate" value="0.0"/>
            </strictClockBranchRates>
        </driftModels>
        <jitter window="1.0 1.0" duplicatesOnly="true">
            <parameter idref="leaf.location"/>
        </jitter>
        <conjugateRootPrior>
            <meanParameter>
                <parameter value="0.0 0.0"/>
            </meanParameter>
            <priorSampleSize>
                <parameter value="0.000001"/>
            </priorSampleSize>
        </conjugateRootPrior>
        <arbitraryBranchRates idref="location.diffusion.branchRates"/>
    </multivariateTraitLikelihood>
    <correlation id="location.correlation" dimension1="1" dimension2="2">
        <matrixParameter idref="location.precision"/>
    </correlation>
    <matrixInverse id="location.varCovar">
        <matrixParameter idref="location.precision"/>
    </matrixInverse>
    <continuousDiffusionStatistic id="location.diffusionRate" greatCircleDistance="true">
        <multivariateTraitLikelihood idref="location.traitLikelihood"/>
    </continuousDiffusionStatistic>

BEAST is crashing because of the driftModels block:
SEVERE: Parsing error - poorly formed BEAST file, sib.complete_cds.directional.xml:
The '<strictClockBranchRates>' element, nested within an element with id, 'location.traitLikelihood', is incorrectly constructed.
The following was expected:
Exactly one ELEMENT of name rate REQUIRED containing
    Exactly one ELEMENT of type Parameter REQUIRED

Am I missing something? Was this option tested in 1.10.4?

Thank you,
Ksenia
Reply all
Reply to author
Forward
0 new messages