Using a starting tree + calibration

[Thijs Janzen]

I'm trying to reconstruct the phylogeny of the Lamprologini (a group of cichlids), using 7 different genes, with each gene having partial coverage of the tribe.

Furthermore I am implementing 4 different calibration points.

So far I have been able to set up everything in BEAUti and get *BEAST to run, however it return an error with a -INF posterior probability for my tmrca's. This suggests that a random starting tree immediately violates the time constraints I have put on some nodes to reflect time calibration data.

As suggested in the tutorials here I have therefore created my own starting tree that does not violate these time constraints, and included this tree via BEAUti in the analysis (import data, and then under trees clicked on "User-specified starting tree").

Unfortunately this does not seem to remedy the problem, as when running *BEAST I still get a -INF posterior probability for all my tmrca's. 

This is all in BEAST 1.7.5 - I have also tried this in BEAST 1.8.2, but somehow *BEAST then transforms my starting tree to a tree with height 150 (whilst my fossil calibrations are on nodes around 5 million years), which also yields -INF on my tmrca priors. 

Or should I enter the starting tree somewhere else?

I have attached my XML file, in case one wants to see things in more detail. The user specified trees are entered around line 3708. 

I did note that the code around "<!-- species starting tree for calibration " (line 4822) is NOT modified by BEAUti, nor does it seem make use of the user specified trees. Could the solution lie here?

Cheers

Thijs

[Thijs Janzen]

Hmm, so far no replies, maybe I have posed my problem a bit too lengthily!

Basically the problem is this - even though I provide pre-cooked trees as prior trees for the gene trees, I still get -Inf posterior probabilities. Is this because I also have to add the newick code of the tree to the coalescent tree prior?

Op woensdag 29 april 2015 15:25:30 UTC+2 schreef Thijs Janzen:

[Nick Matzke]

The first thing to try is to make your node calibration distributions into tight normal distributions -- the normal has no hard edges, so any dates in the starting tree will work.  If this runs, run it for a few million generations to get a "reasonable" starting tree, then put in the calibrations you wanted.

If your starting tree topology disagrees with the topology specified in a node calibration, that will cause this error also, so that's something to check.

Cheers!

Nick

--
You received this message because you are subscribed to the Google Groups "beast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.
To post to this group, send email to beast...@googlegroups.com.
Visit this group at http://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.

[Nick Matzke]

Hi Thijs,

I'm confused, you say it runs now, but you also say you are still getting -INFs?  These seem like contradictory statements.

I'm no expert at starBEAST so I won't be able to help much, but the typical thing on Google Groups is that if want help, you can't leave folks guessing -- if the question is unclear, any answers will be shots in the dark.  Usually it helps to post the chunk of the screen output where the analysis is working, transitioning to the part where you start to see the errors.

Cheers,

Nick

On Mon, Jun 1, 2015 at 9:15 AM, Thijs Janzen <thijs...@gmail.com> wrote:

Allright, I have done that. The tmrca probabilities nicely leveled out in my test run - suggesting that indeed this sample run converged towards trees that are very close to my narrowly defined normal distributions. I have then used the final few trees to build a consensus tree that fits my data.

The error remains however. I have tried using this tutorial to edit my XML to add the starting tree, but this didn't work (as I have multiple genes with multiple substitution models etc).

Other than that I could not find a good tutorial on how to do this for BEAST 1.7.5. 

I have provided my new custom tree to the xml parts associated with the different genes (e.g. <newick id="18S_MAFFT.startingTree" usingDates="true"> custom_newick_code </newick>)

I have not, however, edited the XML concerning the constant population size / anything related with the coalescent model, which I think also needs to be done? I have no idea how though.

Tnx in advance!

Thijs

Op donderdag 7 mei 2015 17:24:08 UTC+2 schreef Nick Matzke:

[Nick Matzke]

Hi -- You can see where the -Inf starts:

==========

Uniform(tmrca(C-lineage))=-Inf, 

    Uniform(tmrca(Congo))=-Inf, 

    Uniform(tmrca(Malawi))=-Inf, 

    Uniform(tmrca(Lamprologini))=-Inf, 

==========

So -- same problem as before. Same solution: switch those priors from Uniform to Normal for an initial run to get a starting tree. Perhaps you did the initial run to get a starting tree, and are now trying to switch back to uniform priors?

If so, either (a) your tight normal priors weren't tight enough the first time, or (b) something about your node topology constraints etc. makes it simply impossible for any tree to satisfy both your topology constraints and your uniform date constraints.

Look very very very carefully at your definition of the clades which are getting node date priors, look at the date priors for those nodes, and look at your starting tree.  Even one species wrongly put in, or left out, of a particular clade constraint could cause the problem, since the date of clade depends on what taxa are in it. This isn't really something other people can solve for you. 

Cheers!

Nick

On Mon, Jun 1, 2015 at 1:51 PM, Thijs Janzen <thijs...@gmail.com> wrote:

Hi Nick,

You are completely right.

So I'm building a tree using 7 different genes, and apply 4 different calibrations. When starting BEAST v1.7.5 using BD.xml (from the previous post), I get the following output (I purged the BEAST contact info at the beginning):

                  BEAST v1.7.5, 2002-2013
Random number seed: 1433180641886
Parsing XML file: BD.xml
  File encoding: MacRoman
Looking for plugins in /Users/janzen/DNA/BEAST3/plugins
Read alignment: alignment1
  Sequences = 103
      Sites = 1316
   Datatype = nucleotide
Read alignment: alignment2
  Sequences = 103
      Sites = 615
   Datatype = nucleotide
Read alignment: alignment3
  Sequences = 103
      Sites = 720
   Datatype = nucleotide
Read alignment: alignment4
  Sequences = 103
      Sites = 1047
   Datatype = nucleotide
Read alignment: alignment5
  Sequences = 103
      Sites = 4552
   Datatype = nucleotide
Read alignment: alignment6
  Sequences = 103
      Sites = 1149
   Datatype = nucleotide
Read alignment: alignment7
  Sequences = 103
      Sites = 797
   Datatype = nucleotide
Site patterns '18S_MAFFT.patterns' created from positions 1-1316 of alignment 'alignment1'
  pattern count = 97
Site patterns '38A_MAFFT.patterns' created from positions 1-615 of alignment 'alignment2'
  pattern count = 143
Site patterns 'COI_MAFFT.patterns' created from positions 1-720 of alignment 'alignment3'
  pattern count = 142
Site patterns 'ND2_MAFFT2.patterns' created from positions 1-1047 of alignment 'alignment4'
  pattern count = 558
Site patterns 'RH1_RENAMED.patterns' created from positions 1-4552 of alignment 'alignment5'
  pattern count = 276
Site patterns 'CYTB_MAFFT2.patterns' created from positions 1-1149 of alignment 'alignment6'
  pattern count = 459
Site patterns 'EDNRB1_TRY.patterns' created from positions 1-797 of alignment 'alignment7'
  pattern count = 267
Creating the tree model, '18S_MAFFT.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Creating the tree model, '38A_MAFFT.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Creating the tree model, 'COI_MAFFT.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Creating the tree model, 'ND2_MAFFT2.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Creating the tree model, 'RH1_RENAMED.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Creating the tree model, 'CYTB_MAFFT2.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Creating the tree model, 'EDNRB1_TRY.treeModel'
  initial tree topology = ((((((((((((((((Lamp_meleagris,Neol_marunguensis),Neol_gracilis),(Chal_cyanophleps,Neol_mondabu)),((Neol_leleupi,Telm_brachygnathus),Neol_christyi)),((Neol_falcicula,Neol_longior),Neol_calliurus)),((((Alto_compressiceps,Lamp_ocellatus),Alto_shell),(Lamp_speciosus,Neol_meeli)),(((Neol_hecqui,Neol_petricola),Neol_brevis),Neol_niger))),((Neol_boulengeri,Neol_pulcher),(Neol_caudopunctatus,Neol_wauthioni))),(((((Lamp_lemairii,Telm_vittatus),(Lepi_attenuatus,Lepi_elongatus)),Neol_timidus),Telm_dhonti),(((Neol_eseki,Neol_splendens),Neol_prochilus),Neol_bifasciatus))),((((((Juli_ornatus,Lepi_mimicus),(Neol_obscurus,Neol_sexfasciatus)),Neol_cylindricus),((Lamp_callipterus,Lepi_kendalli),Juli_transcriptus)),(Neol_modestus,Neol_tetracanthus)),((((Chal_ndobhoi,Chal_popelini),(Neol_brichardi,Neol_helianthus)),(Neol_multifasciatus,Neol_similis)),(Lepi_profundicola,Neol_buescheri)))),(((((Lamp_kungweensis,Lamp_laparogramma),Lepi_nkambae),(Neol_tretocephalus,Neol_variostigma)),Juli_regani),((Alto_calvus,Lamp_signatus),(Neol_mustax,Neol_ventralis)))),((((Juli_marlieri,Neol_walteri),Telm_bifrenatus),(Neol_fasciatus,Neol_nigriventris)),(Neol_olivaceous,Neol_toae))),(((Lamp_ornatipinnis,Lepi_cunningtoni),Chal_brichardi),(Neol_furcifer,Telm_temporalis))),((Juli_dickfeldi,Neol_savoryi),Neol_leloupi)),(Lamp_congoensis,Lamp_teugelsi)),Vari_moorii),((((((((Labi_caeruleus,Pseu_tropheops),(Nimb_livingstonii,Nimb_venustus)),Mela_auratus),Asta_calliptera),(Cten_horei,Trop_moorii)),(((Cyph_frontosa,Para_brieni),(Peri_microlepis,Trig_otostigma)),(Bent_tricoti,Limn_auritus))),(Call_macrops,Opht_ventralis)),((((Spat_marlieri,Tang_irsacae),Eret_cyanostictus),Spat_erythrodon),((Bath_leo,Hemi_stenosoma),Boul_microlepis))))
  tree height = 7.390968321
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Using discretized relaxed clock model.
  over sampling = 1
  parametric model = logNormalDistributionModel
   rate categories = 1
Creating state frequencies model '18S_MAFFT.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating HKY substitution model. Initial kappa = 2.0
Creating site model.
  4 category discrete gamma with initial shape = 0.5
Creating state frequencies model '38A_MAFFT.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating HKY substitution model. Initial kappa = 2.0
Creating site model.
  4 category discrete gamma with initial shape = 0.5
Creating state frequencies model 'COI_MAFFT.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating TN93 substitution model. Initial kappa = 2.0,2.0
Creating site model.
  4 category discrete gamma with initial shape = 0.5
  initial proportion of invariant sites = 0.5
Creating state frequencies model 'ND2_MAFFT2.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating TN93 substitution model. Initial kappa = 2.0,2.0
Creating site model.
  4 category discrete gamma with initial shape = 0.5
  initial proportion of invariant sitesUnhandled attribute (typo?) stateTagName in treeLikelihood:18S_MAFFT.treeLikelihood
 = 0.5
Creating state frequencies model 'RH1_RENAMED.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating HKY substitution model. Initial kappa = 2.0
Creating site model.
  4 category discrete gamma with initial shape = 0.5
  initial proportion of invariant sites = 0.5
Creating state frequencies model 'CYTB_MAFFT2.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating TN93 substitution model. Initial kappa = 2.0,2.0
Creating site model.
  4 category discrete gamma with initial shape = 0.5
  initial proportion of invariant sites = 0.5
Creating state frequencies model 'EDNRB1_TRY.frequencies': Initial frequencies = {0.25, 0.25, 0.25, 0.25}
Creating HKY substitution model. Initial kappa = 2.0
Creating site model.
  initial proportion of invariant sites = 0.5
Loading native NucleotideLikelihoodCore successfully from /Users/janzen/DNA/BEAST3
TreeLikelihood(18S_MAFFT.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 97 unique site patterns.
Branch rate model used: discretizedBranchRates
TreeLikelihood(38A_MAFFT.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 143 unique site patterns.
Branch rate model used: discretizedBranchRates
TreeLikelihood(COI_MAFFT.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 142 unique site patterns.
Branch rate model used: discretizedBranchRates
TreeLikelihood(ND2_MAFFT2.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 558 unique site patterns.
Branch rate model used: discretizedBranchRates
TreeLikelihood(RH1_RENAMED.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 276 unique site patterns.
Branch rate model used: discretizedBranchRates
TreeLikelihood(CYTB_MAFFT2.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 459 unique site patterns.
Branch rate modeUnhandled attribute (typo?) stateTagName in treeLikelihood:38A_MAFFT.treeLikelihood
Unhandled attribute (typo?) stateTagName in treeLikelihood:COI_MAFFT.treeLikelihood
Unhandled attribute (typo?) stateTagName in treeLikelihood:ND2_MAFFT2.treeLikelihood
Unhandled attribute (typo?) stateTagName in treeLikelihood:RH1_RENAMED.treeLikelihood
Unhandled attribute (typo?) stateTagName in treeLikelihood:CYTB_MAFFT2.treeLikelihood
l used: discretizedBranchRates
TreeLikelihood(EDNRB1_TRY.treeModel) using native nucleotide likelihood core
  Ignoring ambiguities in tree likelihood.
  With 267 unique site patterns.
Branch rate model used: discretizedBranchRates
Unhandled attribute (typo?) stateTagName in treeLikelihood:EDNRB1_TRY.treeLikelihood
Using birth-death model on tree: Gernhard T (2008) J Theor Biol, Volume 253, Issue 4, Pages 769-778 In press
Creating swap operator for parameter 18S_MAFFT.branchRates.categories (weight=10.0)
Creating swap operator for parameter 38A_MAFFT.branchRates.categories (weight=10.0)
Creating swap operator for parameter COI_MAFFT.branchRates.categories (weight=10.0)
Creating swap operator for parameter ND2_MAFFT2.branchRates.categories (weight=10.0)
Creating swap operator for parameter RH1_RENAMED.branchRates.categories (weight=10.0)
Creating swap operator for parameter CYTB_MAFFT2.branchRates.categories (weight=10.0)
Creating swap operator for parameter EDNRB1_TRY.branchRates.categories (weight=10.0)
Likelihood computation is using an auto sizing thread pool.
Creating the MCMC chain:
  chainLength=100000000
  autoOptimize=true
  autoOptimize delayed for 1000000 steps
Error running file: BD.xml
The initial model is invalid because state has a zero probability.
If the log likelihood of the tree is -Inf, his may be because the
initial, random tree is so large that it has an extremely bad
likelihood which is being rounded to zero.
Alternatively, it may be that the product of starting mutation rate
and tree height is extremely small or extremely large. 
Finally, it may be that the initial state is incompatible with
one or more 'hard' constraints (on monophyly or bounds on parameter
values. This will result in Priors with zero probability.
The individual components of the posterior are as follows:
The initial posterior is zero: 
  CompoundLikelihood(compoundModel)=(
    MultiSpeciesCoalescent(speciesTree[sptree])=-678636.9194, 
    MixedDistributionLikelihood(MixedDistributions)=-2457.1566, 
    SpeciationLikelihood(speciationLikelihood[speciation.likelihood])=69.1782, 
    BooleanLikelihood=0.0, 
    Uniform(tmrca(C-lineage))=-Inf, 
    Uniform(tmrca(Congo))=-Inf, 
    Uniform(tmrca(Malawi))=-Inf, 
    Uniform(tmrca(Lamprologini))=-Inf, 
    LogNormal(18S_MAFFT.kappa)=-1.8654, 
    Uniform(18S_MAFFT.frequencies)=0.0, 
    Exponential(18S_MAFFT.alpha)=-0.3069, 
    LogNormal(38A_MAFFT.kappa)=-1.8654, 
    Uniform(38A_MAFFT.frequencies)=0.0, 
    Exponential(38A_MAFFT.alpha)=-0.3069, 
    LogNormal(COI_MAFFT.kappa1)=-1.8654, 
    LogNormal(COI_MAFFT.kappa2)=-1.8654, 
    Uniform(COI_MAFFT.frequencies)=0.0, 
    Exponential(COI_MAFFT.alpha)=-0.3069, 
    Uniform(COI_MAFFT.pInv)=0.0, 
    LogNormal(ND2_MAFFT2.kappa1)=-1.8654, 
    LogNormal(ND2_MAFFT2.kappa2)=-1.8654, 
    Uniform(ND2_MAFFT2.frequencies)=0.0, 
    Exponential(ND2_MAFFT2.alpha)=-0.3069, 
    Uniform(ND2_MAFFT2.pInv)=0.0, 
    LogNormal(RH1_RENAMED.kappa)=-1.8654, 
    Uniform(RH1_RENAMED.frequencies)=0.0, 
    Exponential(RH1_RENAMED.alpha)=-0.3069, 
    Uniform(RH1_RENAMED.pInv)=0.0, 
    LogNormal(CYTB_MAFFT2.kappa1)=-1.8654, 
    LogNormal(CYTB_MAFFT2.kappa2)=-1.8654, 
    Uniform(CYTB_MAFFT2.frequencies)=0.0, 
    Exponential(CYTB_MAFFT2.alpha)=-0.3069, 
    Uniform(CYTB_MAFFT2.pInv)=0.0, 
    LogNormal(EDNRB1_TRY.kappa)=-1.8654, 
    Uniform(EDNRB1_TRY.frequencies)=0.0, 
    Uniform(EDNRB1_TRY.pInv)=0.0, 
    Exponential(18S_MAFFT.ucld.stdev)=0.0986, 
    Exponential(38A_MAFFT.ucld.stdev)=0.0986, 
    Exponential(38A_MAFFT.ucld.mean)=-17.0043, 
    Exponential(COI_MAFFT.ucld.stdev)=0.0986, 
    Exponential(COI_MAFFT.ucld.mean)=-17.0043, 
    Exponential(ND2_MAFFT2.ucld.stdev)=0.0986, 
    Exponential(ND2_MAFFT2.ucld.mean)=-17.0043, 
    Exponential(RH1_RENAMED.ucld.stdev)=0.0986, 
    Exponential(RH1_RENAMED.ucld.mean)=-17.0043, 
    Exponential(CYTB_MAFFT2.ucld.stdev)=0.0986, 
    Exponential(CYTB_MAFFT2.ucld.mean)=-17.0043, 
    Exponential(EDNRB1_TRY.ucld.stdev)=0.0986, 
    Exponential(EDNRB1_TRY.ucld.mean)=-17.0043, 
    OneOnX(species.popMean)=0.0, 
    OneOnX(species.birthDeath.meanGrowthRate)=0.0, 
    Uniform(species.birthDeath.relativeDeathRate)=0.0
    Total = -Infinity
  ), 
  CompoundLikelihood(compoundModel)=(
    TreeLikelihood(treeLikelihood[18S_MAFFT.treeLikelihood])=-4019.1938, 
    TreeLikelihood(treeLikelihood[38A_MAFFT.treeLikelihood])=-3487.1814, 
    TreeLikelihood(treeLikelihood[COI_MAFFT.treeLikelihood])=-2681.6767, 
    TreeLikelihood(treeLikelihood[ND2_MAFFT2.treeLikelihood])=-24561.7908, 
    TreeLikelihood(treeLikelihood[RH1_RENAMED.treeLikelihood])=-11059.9335, 
    TreeLikelihood(treeLikelihood[CYTB_MAFFT2.treeLikelihood])=-11146.2114, 
    TreeLikelihood(treeLikelihood[EDNRB1_TRY.treeLikelihood])=-10667.0004
    Total = -67622.9880369571
  )
  Total = -Infinity
For more information go to <http://beast.bio.ed.ac.uk/>.
BEAST has terminated with an error. Please select QUIT from the menu.

Hence, the total likelihood is -Infinity, which is caused by:

    Uniform(tmrca(C-lineage))=-Inf, 

    Uniform(tmrca(Congo))=-Inf, 

    Uniform(tmrca(Malawi))=-Inf, 

    Uniform(tmrca(Lamprologini))=-Inf, 

I have supplied a user defined tree in the xml file. I have given the same tree as initial tree for each gene, but have not edited the coalescent part (as suggested in  this tutorial, however I don't see how that should work with starBeast)

I hope that clarifies things.

Thijs

 

Op maandag 1 juni 2015 18:43:54 UTC+2 schreef Nick Matzke:

[Thijs Janzen]

Hi Thijs,

[Thijs Janzen]

Allright, I have done that. The tmrca probabilities nicely leveled out in my test run - suggesting that indeed this sample run converged towards trees that are very close to my narrowly defined normal distributions. I have then used the final few trees to build a consensus tree that fits my data.

The error remains however. I have tried using this tutorial to edit my XML to add the starting tree, but this didn't work (as I have multiple genes with multiple substitution models etc).

Other than that I could not find a good tutorial on how to do this for BEAST 1.7.5. 

I have provided my new custom tree to the xml parts associated with the different genes (e.g. <newick id="18S_MAFFT.startingTree" usingDates="true"> custom_newick_code </newick>)

I have not, however, edited the XML concerning the constant population size / anything related with the coalescent model, which I think also needs to be done? I have no idea how though.

Tnx in advance!

Thijs

Op donderdag 7 mei 2015 17:24:08 UTC+2 schreef Nick Matzke:

The first thing to try is to make your node calibration distributions into tight normal distributions -- the normal has no hard edges, so any dates in the starting tree will work.  If this runs, run it for a few million generations to get a "reasonable" starting tree, then put in the calibrations you wanted.

[Thijs Janzen]

Allright, I have set everything up in BEAST v2 this time - and using the tutorial on how to provide a starting tree I was able to get everything running (hence, the tree is fine).

Then, as a last check I used that tree again in my BEAST v1.7.5 XML file, and this again did not run. So whatever it was, it was not the tree...

It is now very happily running on BEAST v2, so my troubles are gone!

Op dinsdag 2 juni 2015 07:44:47 UTC+2 schreef Thijs Janzen: