veg / hyphy

HyPhy: Hypothesis testing using Phylogenies
http://www.hyphy.org
Other
211 stars 69 forks source link

MEME errorcode1 repeated message issues #1692

Closed cmariencheck closed 7 months ago

cmariencheck commented 7 months ago

Hi,

I have been trying to run MEME with several different files (see attached) and some files (TLR7; TLR8) will be continuously rejected with the attached error code message ( MPI_ABORT was invoked on rank 0 in communicator MPI_COMM_WORLD with errorcode 1; PMIX ERROR: UNREACHABLE in file server/pmix_server.c at line 1741). Any recommendations on changes to help avoid these errors? Attached is another file I made that was accepted with no issues.

Thank you! Clara MSL3_tree_aligned.txt TLR7_tree_combined_file.txt TLR8_tree_sequence_combined.txt error_log_TLR8.txt error_log_TLR7.txt

spond commented 7 months ago

Dear @cmariencheck,

The issue appears to be caused by the mismatch between the tree/taxa list of the NEXUS file and the sequences actually present. For example, in TLR7, the tree and the TAXA block refer to Pan_troglodytes_verus while the MATRIX block refers to Pan_troglodytes. HyPhy gives precedence to sequence names from the MATRIX block if present. Simply editing those (Pan_troglodytes -> Pan_troglodytes_verus and Cercocebus_torquatus -> Cercocebus_torquatus_atys) resolves the issue.

Similar issues affect TLR8.

Best, Sergei

TLR7_tree_combined_file.txt

cmariencheck commented 7 months ago

Hi @spond

Thank you for your response! I see the issue with the TLR7 file but those were already remedied with the TLR8 file. Is there something else wrong with it?

Best,

Clara

cmariencheck commented 7 months ago

Hi @spond

The issue is now arising with an expanded format of the MSL3 file where I just added 2 more taxa. It is attached, and I am getting the same error message. See attached files. Please advise.

Clara [MSL3_tree_seq_combined.txt](https://github.com/veg/hyphy/files/14465093/MSL3_tree_seq_combined.txt) [MSL3_error.txt](https://github.com/veg/hyphy/files/14465095/MSL3_error.txt)

spond commented 7 months ago

Dear @cmariencheck,

Right, I see. This is not an issue with your data. It's a numerical instablity warning in HyPhy

@stevenweaver : could we enable ENV="TOLERATE_NUMERICAL_ERRORS=1;" by default on all datamonkey analyses?

Best, Sergei

PS The results are here: https://observablehq.com/@spond/meme?url=https://dl.dropbox.com/scl/fi/t30go41lfhbn75qdjq3c7/MSL3_tree_seq_combined.txt.MEME.json?rlkey=z73yxbtpi4wox8e1m60ih9cvt

stevenweaver commented 7 months ago

Completed