delimitation - Pas-Kapli/tutorials GitHub Wiki

1.4. Run ``species'' delimitation with mptp:

Check the mptp options:

$ mptp --help

Run delimitation with the multi-rate algorithm

If the phylogeny is unrooted use the ``--outgroup taxon_name1,taxon_name2'' for rooting the phylogeny prior to the delimitation

$ mptp --ml --multi --tree_file RAxML_bestTree.Anolis --output_file Anolis_multi_delimitation

Check the output files:

$ firefox Anolis_multi_delimitation.svg

$ nano Anolis_multi_delimitation.txt

Q: How many putative species are delimited?

Detect minimum branch length:

$ mptp --tree_file RAxML_bestTree.Anolis --minbr_auto Anolis_mafft.fas.reduced.fasta --output_file minbr

Re-run delimitation taking the minimum branch length into account:

$ mptp --ml --multi --tree_file RAxML_bestTree.Anolis --output_file Anolis_multi_delimitation --minbr 0.0014695325

Re-run but crop the outgroup taxa:

$ mptp --ml --single --tree_file RAxML_bestTree.Anolis --output_file signle --outgroup GBGC1209413_Polychrus --outgroup_crop

Q: Does the delimitation change when by using the minimum branch length or the crop-outgroup command? Why? When would these options matter?

Repeat the exercise with the single-rate algorithm:

Q: How do the results differ from the multi-rate PTP?

1.5. Support values:

Perform MCMC sampling (proportional to the likelihood) for the multi and the single rate PTP: $ mptp --tree_file RAxML_bestTree.Anolis --minbr 0.0014695325 --mcmc 10000000 --mcmc_log 1000000 --multi --mcmc_runs 2 --output_file mcmc_multi

$ firefox mcmc_multi.SOME_NUMBER.combined.svg

Q: Did the two runs converge? Q: Did the runs converge to the Maximum Likelihood solution?