epa-ng

[Sigmund Jensen]

Hi Placement group,

epa-ng installed with miniconda3 worked fine until it stopped working. Upon execution it now just says 'Aborted¨. If anyone know, may I please have a suggested solution?

Regards, Sigmund

[Lucas Czech]

Hi Sigmund,

without further information, that is hard to tell. What command did you use exactly? What type of machine are you running it on? Was there any other output prior to the abort? How large is your dataset, and did you check that EPA-ng is not just simply running out of memory?

Please post some more details here ;-)

Cheers and so long
Lucas

--
You received this message because you are subscribed to the Google Groups "Phylogenetic Placement" group.
To unsubscribe from this group and stop receiving emails from it, send an email to phylogenetic-plac...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/phylogenetic-placement/973a7ca0-c5b3-4e6f-a016-e3ca9434b611n%40googlegroups.com.

[Sigmund Jensen]

Hi Lucas,

many thanks for the reply. I used these commands:

epa-ng --ref-msa $REF_MSA --tree $TREE --query $QRY_MSA --model $MODEL.

I am running it on a HP Spectre x360 (just a standard PC with RAM 8 GB and Windows 10 ) using Windows Subsystem Linux.

I can't remember any other output prior to the abort (I did not save a log of screen output). My dataset was all the Silva Nitrosococcales 16S rRNA gene sequences about 200 sequences. It worked fine with a smaller dataset of about 1/10 the size. My PC could be running out of memory (I do not know how to check that). Running iqtree2 with an alignment (nucleotide positions with only gaps removed) of all the Gammaproteobacteira (about 30 000 sequences) iqtree2 stopped and I think because of lack of memory. Therefore I made the tree with the lower level taxon (fewer sequences). Apart from memory I wonder if it might have to do with the --model (I used JC to keep it simple).

What I am trying to do is to find relatives of my short sequences by placing them into a tree of Silva sequences (as an alternative to ARB). Then pick relatives for a backbone tree and add the short sequences to this tree, for a manuscript.

Regards, Sigmund

[Pierre Barbera]

Hi Sigmund,

the "Aborted" error is usually due to memory usage; you can check the task manager (or the top  or htop command from inside your WSL) while it's running and see how high it gets. Very likely it's a memory issue though, but this doesn't usually happen with reference trees that small. How wide is the alignment? How many sequences are you trying to place?

Sidenote: It's not clear from what you write how you fed in the model, the correct way is to pass either a RAxML-8 .info file, a raxml-ng model file, or from the iqtree report file (that one is the least tested). epa-ng will not re-optimize the model you give it, so if you just specified --model JC it will have taken default parameters.

Pierre

To view this discussion on the web, visit https://groups.google.com/d/msgid/phylogenetic-placement/68128478-1325-48d5-9286-d319886268f5n%40googlegroups.com.

[Sigmund Jensen]

Hi Pierre,

thanks for replying although my question may not provide much phylo-programming insight. That 'Aborted' error disapeared when I deleted the file 'epa_info_log'. Now its working. I just have to delete or rename that file prior to a new run. My alignment is about 1700 positions wide (i.e. about 1500 nucleotide positions 16S rRNA gene plus gaps from the alignment). So far I have placed 3 sequences per tree. The largest alignment I have placed sequences into, so far, contains about 1400 sequences). I have not tried that RAxML-8 yet. I am trying to keep it simple and run both iqtree2 and epa-ng with model HKY (I have used that model before with ARB).

Regards, Sigmund

[Lucas Czech]

Interesting.

Pierre, could it be that it aborts when the file is present, because it does not want to overwrite - but for some reason, instead of printing the exception message, it just aborts? I've had another user of some tool saying that as well recently as well - exceptions that are only printed in some cases, but not always. Maybe it makes sense then to have a global high-level catch, print to stdout and log file, and then re-throw?

Sigmund, does that mean things work now? If there is any more trouble, let us know!

Cheers and so long
Lucas

To view this discussion on the web, visit https://groups.google.com/d/msgid/phylogenetic-placement/ca227b72-67e5-4c36-aaf5-77368617f726n%40googlegroups.com.

[Sigmund Jensen]

Hi Lucas,

yes, it means things work now. I just have to remove or rename that file 'epa_info.log' prior to a new run. Yes, I let you know of any more trouble and thanks for help this far.

Regards, Sigmund