Assertion failed: function pllEvaluateLikelihood

[biomendi]

Hi Diego,

I have a 40 sequence alignment of length 117,579 derived from capture data (lots of missing data), a partition file with 84 subsets (computed with partitionfinder), and an ML tree inferred with fasttree. The tree has the long branch issue, which I would like to fix using ForeSeqs.

I installed ForeSeqs on my mac server and run successfully the test data. However, when I tried to use it with my data, the program did run for while (until partition 64) and then failed.

Predicting subtree
  - Estimated branch length for node 2 (partition Subset64) = 0.02462
    - Substitutions from ancestral = 0.77%
Tree with stolen branches for partition 63:
.
.
.
Per site Gamma rate categories assignment:
  0: 0.01540 (81.63%)
  1: 0.17479 (0.70%)
  2: 0.72104 (0.00%)
  3: 3.08878 (17.67%)

Predicting subtree
  - Estimated branch length for node 23 (partition Subset65) = 0.02351
    - Substitutions from ancestral = 3.14%
Tree with stolen branches for partition 64:
.
.
.
Assertion failed: (PLL_ABS(tr->likelihood - likelihood) < 0.00001), function pllEvaluateLikelihood, file evaluateGenericSpecial.c, line 1985.
Abort trap: 6

Do you have any insights about this error?

If you need a copy of the input files, please let me know.

Thanks a lot.

[ddarriba]

Hi Santiago,

Could you please send me the input file to my personal email? ddarriba (at) udc.es. It will be much faster if I can check it locally.

Best, Diego

[biomendi]

Hi Diego,

Input files should now be in your inbox.

Thanks again and I will be waiting for your answer.

[scastanedarico]

Hi Diego,

I'm very interested in how did you fix Santiago's problem because I have the same one after partition 22:

Assertion failed: (PLL_ABS(tr->likelihood - likelihood) < 0.00001), function pllEvaluateLikelihood, file evaluateGenericSpecial.c, line 1985. Abort trap: 6

Thank you so much!

Best, Susette