Help for "IQtree crashes with signal aborted"

[Ma B]

Hi,

First thank you for developping IQtree :)

I am running into an issue and have been trying several things.

Here is my command line, it works well for an alignment file of 2000 sequences as an input but crashes with my final file of 86k sequences. I have both been trying it on my local computer and on our cluster.

Command line: iqtree -s <path to fasta file> -m K3Pu+F+G4 -bo 10
Seed:    815072 (Using SPRNG - Scalable Parallel Random Number Generator)

Part of the log with error (for the run on my local computer):

WARNING: 86622 sequences contain more than 50% gaps/ambiguity
****  TOTAL   56.00%  18 sequences failed composition chi2 test (p-value<5%; df=3)

===> START BOOTSTRAP REPLICATE NUMBER 1

Creating bootstrap alignment (seed: 815072)...

Create initial parsimony tree by phylogenetic likelihood library (PLL)... 2693.687 seconds

NOTE: 2112 MB RAM (2 GB) is required!
Estimate model parameters (epsilon = 0.100)
1. Initial log-likelihood: -325369.792
2. Current log-likelihood: -315330.127
3. Current log-likelihood: -314971.524
4. Current log-likelihood: -314969.954
Optimal log-likelihood: -314969.918
Rate parameters:  A-C: 1.00000  A-G: 1.40015  A-T: 0.52119  C-G: 0.52119  C-T: 1.40015  G-T: 1.00000
Base frequencies:  A: 0.254  C: 0.222  G: 0.232  T: 0.292
Gamma shape alpha: 0.845
Parameters optimization took 4 rounds (3081.983 sec)
Computing ML distances based on estimated model parameters...
ERROR: STACK TRACE FOR DEBUGGING:
ERROR: 1   funcAbort()
ERROR: 2   ()
ERROR: 3   gsignal()
ERROR: 4   abort()
ERROR: 5   __gnu_cxx::__verbose_terminate_handler()
ERROR: 6   ()
ERROR: 7   ()
ERROR: 8   __cxa_rethrow()
ERROR: 9   operator new(unsigned long)
ERROR: 10   PhyloTree::computeDist(Params&, Alignment*, double*&, double*&)
ERROR: 11   computeMLDist(Params&, IQTree&, double, double)
ERROR: 12   runTreeReconstruction(Params&, IQTree*&)
ERROR: 13   runStandardBootstrap(Params&, Alignment*, IQTree*)
ERROR: 14   runPhyloAnalysis(Params&, Checkpoint*)
ERROR: 15   main()
ERROR: 16   __libc_start_main()
ERROR: 17   ()
ERROR:
ERROR: *** IQ-TREE CRASHES WITH SIGNAL ABORTED
ERROR: *** For bug report please send to developers:
ERROR: ***    Log file: Documents/TARACommunityDistributionNAP/WORK/rDNA_V9_18S_ribotypes_diatoms/Phylogeny/Phylogeny_Sina_ExaML_Diatoms_SO/diatoms_aligned_trimmed.fasta.log
ERROR: ***    Alignment files (if possible)

And here is the error part when launching it on the cluster:

WARNING: 86622 sequences contain more than 50% gaps/ambiguity
****  TOTAL   56.00%  18 sequences failed composition chi2 test (p-value<5%; df=3)

===> START BOOTSTRAP REPLICATE NUMBER 1

Creating bootstrap alignment (seed: 600398)...

Create initial parsimony tree by phylogenetic likelihood library (PLL)... 8042.695 seconds

NOTE: 1977 MB RAM (1 GB) is required!
Estimate model parameters (epsilon = 0.100)
1. Initial log-likelihood: -418723.320
2. Current log-likelihood: -329717.488
3. Current log-likelihood: -329667.347
4. Current log-likelihood: -329665.307
5. Current log-likelihood: -329664.876
6. Current log-likelihood: -329664.761
Optimal log-likelihood: -329664.721
Rate parameters:  A-C: 1.00000  A-G: 1.50455  A-T: 0.65018  C-G: 0.65018  C-T: 1.50455  G-T: 1.00000
Base frequencies:  A: 0.227  C: 0.148  G: 0.383  T: 0.241
Gamma shape alpha: 1.762
Parameters optimization took 6 rounds (20665.391 sec)
Computing ML distances based on estimated model parameters...ERROR: STACK TRACE FOR DEBUGGING:
ERROR:
ERROR: *** IQ-TREE CRASHES WITH SIGNAL ABORTED
ERROR: *** For bug report please send to developers:
ERROR: ***    Log file: output_data/run-37981/2021-07-02-diatoms_aligned_trimmed.fasta.MFP.0.log
ERROR: ***    Alignment files (if possible)

On the cluster when it crashes the memory usage is around 3G but more is allocated.

On my PC I have 30G of RAM.

Do you know where this could be coming from?

Let me know if you need additionnal info, ie full log file or alignment file.

Best,

Mathilde

[Minh Bui]

Hi Mathilde,

Which IQ-TREE version did you use?

Older versions cannot handle >65K sequences at the step of constructing the distance matrix, that this crash message printed. You can you to rerun with additional option -t PARS to avoid computing the distance trees.

I believe later version resolved this issue.

Cheers,

Minh

--
You received this message because you are subscribed to the Google Groups "IQ-TREE" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iqtree+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/e162be8e-0b82-4b91-a5eb-4c448842bd87n%40googlegroups.com.

[Ma B]

Hi Minh,

Thank you very much for your reply.

We have been using the latest version both on my PC and the cluster.

We tried on an other cluster but get these messages (we asked for 16G of memory):

WITH bootsrap:

===> START BOOTSTRAP REPLICATE NUMBER 1

Creating bootstrap alignment (seed: 54712)...

CHECKPOINT: Initial tree restored

NOTE: 2022 MB RAM (1 GB) is required!

NOTE: 4 processes are running on the same host r1i0n11

Measuring multi-threading efficiency up to 0 CPU cores

BEST NUMBER OF THREADS: 1

CHECKPOINT: Model parameters restored, LogL: -1260872.572

Computing ML distances based on estimated model parameters...

slurmstepd: error: Detected 1 oom-kill event(s) in step 621496.0 cgroup. Some of your processes may have been killed by the cgroup out-of-memory handler.

srun: error: r1i0n11: task 0: Out Of Memory

srun: Terminating job step 621496.0

slurmstepd: error: *** STEP 621496.0 ON r1i0n11 CANCELLED AT 2021-07-31T13:18:03 ***

slurmstepd: error: Detected 1 oom-kill event(s) in step 621496.batch cgroup. Some of your processes may have been killed by the cgroup out-of-memory handler.

WITHOUT bootstrap:

Create initial parsimony tree by phylogenetic likelihood library (PLL)... 3933.792 seconds

NOTE: 3145 MB RAM (3 GB) is required!

NOTE: 2 processes are running on the same host r1i0n9

Measuring multi-threading efficiency up to 0 CPU cores

BEST NUMBER OF THREADS: 1

Estimate model parameters (epsilon = 0.100)

1. Initial log-likelihood: -642440.945

2. Current log-likelihood: -632227.563

3. Current log-likelihood: -631224.129

4. Current log-likelihood: -630669.471

5. Current log-likelihood: -630483.681

6. Current log-likelihood: -630103.913

7. Current log-likelihood: -630057.642

8. Current log-likelihood: -630057.125

slurmstepd: error: Detected 2 oom-kill event(s) in step 606709.0 cgroup. Some of your processes may have been killed by the cgroup out-of-memory handler.

srun: error: r1i0n9: task 1: Out Of Memory

srun: Terminating job step 606709.0

slurmstepd: error: *** STEP 606709.0 ON r1i0n9 CANCELLED AT 2021-07-30T05:59:43 ***

slurmstepd: error: Detected 2 oom-kill event(s) in step 606709.batch cgroup. Some of your processes may have been killed by the cgroup out-of-memory handler.

I hope that can help resolve the issue?

Best,

Mathilde

[Minh Bui]

Hi Mathilde,

As I said earlier, iqtree v1 cannot handle >65K sequences.

Have you tried v2? Alternatively use v1 but with -t PARS option, to avoid computing the distance tree.

Cheers

Minh

To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/a9dd67cc-75b4-4fb1-a3f7-f731051410d7n%40googlegroups.com.

[Ma B]

Hi Minh,

Thank you very much, yes I have been using the CoVid release COVID-19 release 2.1.3.

Best,

Mathilde

[Minh Bui]

And did it solve the problem?

Minh

To view this discussion on the web visit https://groups.google.com/d/msgid/iqtree/11c43e8a-20d6-445d-b99d-f397e04533cen%40googlegroups.com.

[Ma B]

Hi,

Sorry for the delay I was out of office.

No, it is the same result.

Best,

Mathilde

[Ma B]

I can try to downsize the number of sequences to 65k, if that could help? Could it be something else since the latest version should handle more than 65k sequences as you said?

Thank you,

Mathilde