Mugsy multiple alignment and core dumped

Question

[vasilislenis]

Hello,

I am trying to make a reference free multiple alignment of 8 fly genomes with mugsy. Mugsy was running for about a week and everything went well. It created the maf files based on each genome and finally it crashed by giving a core dumped.

I checked the memory and the upper limit that it used, and it was 50G. I had allocated 100G, so it wasn't a memory issue.

It crashed when it tried to construct the mugsywga.out file. Is there anyone who had a similar issue and somehow he solved it?

Also, is it possible to continue the job from the checkpoint that it stopped?

Thank you very much in advance,

Vasilis.

[GenoMax]

Sort of obvious but you did have write permissions at the location where the output was being written to and there was sufficient space available?

Answer 1

[kloetzl]

Also, is it possible to continue the job from the checkpoint that it stopped?

Yes that is possible. Mugsy is simply a perl wrapper around a bunch of other tools. If you have all the required files you can manually run the last commands again.

[vasilislenis]

Thank you very much for your reply. It crashed when it was trying to run mugsyWGA. The last command that was trying to run was:

mugsyWGA --outfile ~/tse_tse_pseudo --seq ~/tse-tse/all/tse_tse_pseudo.all.fsa --aln ~/tse-tse/all/tse_tse_pseudo.xmfa --distance 1000 --minlength 30 --unique true > ~/tse-tse/all/tse_tse_pseudo.mugsywga.out 2> ~/tse-tse/all/tse_tse_pseudo.mugsywga.stderr

The tse_tse_pseudo.mugsywga.out was empty and the stderr at the end said that it was building an alignment graph from 36796296 matches.

So, if I run the previous command as it is, will it be ok?

Thank you very much for one more time.

[kloetzl]

So, if I run the previous command as it is, will it be ok?

I cannot promise anything. All I can say is that you do not have to rerun the pairwise nucmer comparison again. Good luck.

[vasilislenis]

Thanks again for your reply. Unfortunately, it didn't work. For some reason that I cannot understand it fills only the mugsywga.stderr by saying that it builds the graph.

If you have any other suggestion fo whole genome multiple alignment I would really appreciate it.

[kloetzl]

Unfortunately 8 fly genomes is too much data for any other tool I know of. Sorry. ☹