"A program that Partitionfinder uses failed." PF2 stops analyzing!

[Johanna Dechert]

Hello everyone,

I used PartitionFinder 2.1.1 for a Phylip file of plastome sequences. It contains only extracted genes aligned and concatenated with "Nylander". Unfortunately PF2 stopped analyzing after a couple of hours with the following error message:

INFO     | 2020-05-27 17:18:35,717 |       Finished subset 6015/6216, 96.77 percent done

Exception in thread Thread-24:

Traceback (most recent call last):

  File "/usr/lib/python2.7/threading.py", line 801, in __bootstrap_inner

    self.run()

  File "/mnt/work7/johanna/Masterarbeit/partitionfinder-2.1.1/partfinder/threadpool.py", line 115, in run

    cmd, args = self.pool.next_task()

  File "/mnt/work7/johanna/Masterarbeit/partitionfinder-2.1.1/partfinder/threadpool.py", line 81, in next_task

    task = self.tasks[self.curtask]

IndexError: list index out of range

ERROR    | 2020-05-27 17:18:35,732 | A program that Partitionfinder uses failed. Output follows, in case it's helpful for finding the problem

ERROR    | 2020-05-27 17:18:35,732 |

ERROR    | 2020-05-27 17:18:35,732 | .......................... ooooooooooooooooooooooooooooo CURRENT SETTINGS ooooooooooooooooooooooooooooooooooo .......................... . Sequence filename: e1aaf30b6c82b98fd30b318b185a5324.phy . Data type: dna . Alphabet size: 4 . Sequence format: interleaved . Number of data sets: 1 . Nb of bootstrapped data sets: 0 . Model name: Custom . Proportion of invariable sites: 0.000000 . Number of subst. rate categs: 4 . Gamma distribution parameter: estimated . 'Middle' of each rate class: mean . Nucleotide equilibrium frequencies: user-defined . Optimise tree topology: no . Evaluated tree: file "user tree (filtered_source.phy_phyml_tree.txt)" . Optimise branch lengths: yes . Optimise substitution model parameters: yes . Run ID: K81+G . Random seed: 1590592712 . Subtree patterns aliasing: no . Version: 20120412 oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo . 17 patterns found (out of a total of 167 sites). . 161 sites without polymorphism (96.41%). . Reading tree... . ( 0 sec) [ -281.7077] [Branch lengths ] . ( 0 sec) [ -281.1305] [GTR parameters ] . ( 0 sec) [ -281.0999] [Alpha ][ 1.618034] . ( 0 sec) [ -281.0999] [Branch lengths ] . ( 0 sec) [ -281.0999] [GTR parameters ] . ( 0 sec) [ -281.0792] [Alpha ][ 2.618034] . ( 0 sec) [ -281.0792] [Branch lengths ] . Eigenvalues/vectors computation does not converge : computation cancelled . Type enter to exit.

I really couldn´t find the reason for this error. I used PF2 before with a similar input file and it worked fine.

I attached the subset file, which was analyzed when PF2 failed.

Hopefully someone can make more sense out of this than I can?!

Thanks a lot in advance and stay healthy!!!!!!!!!!

Johanna

[Rob Lanfear]

Hi Johanna,

This seems to be caused by PhyML being unable to successfully analyse some very low-diversity subsets.

I'd recommend trying a few things:

1. Use `--raxml` to use raxml instead. the Final partitioning scheme will be very similar to what you would get with PhyML, but raxml may be more successful in analysing the low-diversity seuqences.
2. Try ModelFinder in IQ-TREE2. It can also do the greedy analysis, and we are in the process of finalising a manuscript on ModelFinder 2, which will essentially replace PartitionFinder2. Details are here: http://www.iqtree.org/doc/Advanced-Tutorial#choosing-the-right-partitioning-scheme

Rob