Euler-SR de novo assembly error during runtime

[allenyu]

Hi all,

I have encountered terminate called after throwing an instance of 'std::bad_alloc' message during euler-sr assembly in the assemblesec.pl stage. Does anybody have experienced that before? Thanks

${EUSRC}/assembly/Assemble.pl sample1_reads_cat_qualfil.fa -ruleFile sample1_reads_cat_qualfil.rules

mkdir -p fixed; /home/allenyu/source/euler-sr.1.1.2/assembly/x86_64/readsToSpectrum sample1_reads_cat_qualfil.fa -ruleFile fixed/sample1_reads_cat_qualfil.fa.spect -printCount
/home/allenyu/source/euler-sr.1.1.2/assembly/x86_64/estErrorDist fixed/sample1_reads_cat_qualfil.fa.spect -binary
/home/allenyu/source/euler-sr.1.1.2/assembly/x86_64/sortIntegralTupleList fixed/sample1_reads_cat_qualfil.fa.spect -printCount -minMult 0
Using minimum multiplicity: 2
/home/allenyu/source/euler-sr.1.1.2/assembly/x86_64/fixErrors sample1_reads_cat_qualfil.fa fixed/sample1_reads_cat_qualfil.fa.spect -ruleFile fixed/sample1_reads_cat_qualfil.fa -minMult 2 -maxScore 3 -startScore 2 -stepScore 1 -minVotes 2 -edgeLimit 3 -replaceN
/home/allenyu/source/euler-sr.1.1.2/assembly/assemblesec.pl fixed/sample1_reads_cat_qualfil.fa -vertexSize -ruleFile
terminate called after throwing an instance of 'std::bad_alloc'
what(): St9bad_alloc

[BaCh]

terminate called after throwing an instance of 'std::bad_alloc' what(): St9bad_alloc

This occurs when the operating system cannot provide memory as requested by the program (euler-sr in this case). It's basically an "out-of-memory" error written the C++ way.

This can be due to:

1. you really have too much data for your machine
2. the program made a miscalculation somewhere and asks for much more memory than it would really need

As to which of these possibilities apply to your situation ... perhaps you might want to ask the author? Don't forget to specify how much memory your machine has and how big the files are that you are trying to process ... that saves a few mails back and forth

Regards,
B.

[allenyu]

Thanks for your reply. My machine is a Mac pro 8 core xeon 3ghz with 32 GB ram.
The paired end solexa reads are 532MB each, with a reference genome size around 4MB. The version of euler-sr is 1.1.2. I guess my computer's resource sufficient?

[mchaisso]

Yes, your resources are sufficient. There are problems getting it to run on a mac, mainly due to the fact that I do not have a mac to test it out on... Glenn Tesler is compiling euler-sr for sparc, and has found a bunch of potential portability conflicts. He should post that soon, and you can try it out to see if it fixes your problems.

best,
-mark

Thanks for your reply. My machine is a Mac pro 8 core xeon 3ghz with 32 GB ram.
The paired end solexa reads are 532MB each, with a reference genome size around 4MB. The version of euler-sr is 1.1.2. I guess my computer's resource sufficient?

[allenyu]

Mark, you didn't run into this problem in your system? May I know the OS you are using?
The OS installed on my mac is CentOS 5.3.

Best,
Allen

[allenyu]

Does anybody know how can I contact the author of euler-sr? I applied an account on their websites for days but the admin still doesn't approve my membership.

[BaCh]

Does anybody know how can I contact the author of euler-sr? I applied an account on their websites for days but the admin still doesn't approve my membership.

"You don't need to contact them, they'll contact you!"

Sorry, couldn't resist to that joke. In fact, Mark (mchaisso) is the author

[ECO]

Does anybody know how can I contact the author of euler-sr? I applied an account on their websites for days but the admin still doesn't approve my membership.

I'll make sure he sees this.

[allenyu]

LOL. Mark, sorry for the ignorance. I would kindly wait for your reply.

[sediener]

Hi,

I've run into a similar error message during the first program in the pipeline: readsToSpectrum

Is there any upper limit to the number of sequences Euler can handle? I am trying to assemble illumina + 454 reads on a Linux box with 256GB or memory.

thanks,
Steve

[mchaisso]

Sorry for the *very* delayed response... my euler-sr time is very limited these days. The website died during a hard drive crash right at the time I was defending, so instead of fixing it, I just put a link on the site to the code itself, and begged people to register later for the code. The project is back up and running now, under different and very talented leadership, so the site should be accepting logins in the near future.

bad_alloc's are typically from either a bug, or allocating too much memory. A bug causing this was recently fixed, but it is taking its time to get into the release due to the shift in man-power.