Newbler 2.5.3 error

[blindtiger454]

has anybody else encountered this error during output phase of Newbler 2.5.3:

Call Stack:
8: safeAssert(char const*, char const*, unsigned int, char const*)
7: Sequence::getCafieScore(int)
6: Alignment::setSnpMode(int)
5: Alignment::set(ChordMatrix*, std::vector<ChordRegion, std::allocator<ChordRegion> > const&, bool, int)
4: OutputPhase::setupAlignmentObject(ScaffoldIterJob*)
3: OutputPhaseWorkerThread::run()
2: Thread::_dispatch(void*)
1: /lib/libpthread.so.0 [0x7f30402983f7]

Error: An internal error has occurred in the computation.
Assertion: pos >= 1 && pos <= getLength()
Location: 'int Sequence::getCafieScore(int)' [Sequence.cpp:259]
Please report this error to your customer support representative.


This error occurs while generating output. Halfway through the "Generating output..." phase, the error pops up. I am trying to assemble cdna, including sff file with sanger reads. My sff file contains 1,285,287 reads and my sanger file contains 1,913,387 reads (it was created from Illumina assembly. I am trying to do hybrid assembly via Newbler). Using latest version of Ubuntu 64bit, 25GB RAM and 8 processors. Command line used is: runAssembley -cdna -cpu 8 infile.sff sangers.fna

I have tried using the -siod option, but get the same error. I have also seen the same error when I only use SFF files. i have successfully assembled single SFF files with ~1,000,000 reads. My first hunch is a memory issue, since the error occurs when I have > ~2,000,000 reads.

[flxlex]

Apart from monitoring the memory usage ('top' command), I think I would follow the suggestion and 'report this error to your customer support representative' ...

(I've had these before and they indicate a crash of newbler, potentially caused by a bug)

[A_Morozov]

I have another question regarding Newbler assembly. Called gsAssembler via GUI with five .sff's, all settings default; computations goes on and probably finishes, but all that I get at end is

Computing signals...
Warning: required disc space (13 GB) exceeds available disc space (1 GB)
Warning: The computation ran out of memory and is unable to continue.

The same result on desktop (Ubuntu 10.04 x32) and on cluster. Obviously, in both cases there is much more than 13 Gigs of free space on discs, so I'm using it wrong way. Would like, where exactly, please.

[flxlex]

Warning: required disc space (13 GB) exceeds available disc space (1 GB)
Warning: The computation ran out of memory and is unable to continue.

It is strange newbler would say it ran out of memory while it seems to have a disc space problem (or at least, thinks it has one). Can you exclude it is in fact not a memory problem? Using 'top', for example? And, check disc space with a shell-script that reports 'df -h' each minute or so?

[A_Morozov]

Yeah, already checked that, all OK. Kinda weird, but sometimes "Large and complex genome" or "Quick output" helps it to finish assembly. Or loading and launching once again crushed project. And sometimes it does not (with dataset and all other settings remaining the same).
By the way, I've already encountered every of this errors alone, both together and neither, so they seem to happen independently of each other.

[blindtiger454]

Has Roche come out with a bug fix for Newbler 2.5.3 yet? The problem I reported in this thread seems to common, especially among Ubuntu users. Emailing customer service (multiple times) has resulted in no reply.

[ckidner]

I get the same thing and am just working my way through repeating the assembly over and over with different options to try and get round it.

[IrisZhu]

I just got exactly the same thing. My input is 0.65M reads with an average length of 370, for a cDNA assembly. Did anybody get a solution?

[IrisZhu]

I separated my 454 input file into several smaller ones and run them in one batch and the output is normal.

[polarb]

I have found that one transcript assembly out of three failed to complete using newbler/2.5.3.

Those that succeeded had raw *.sff sizes 1.4G and 0.8G. The one that failed was much larger 3.2G.

For what it's worth I re-ran the 3.2G files with newbler/2.3 and it completed successfully.

Another thread has been trying to get an answer from roche http://seqanswers.com/forums/showthread.php?t=11303

[karenchu]

so more than 3.2G files fail?


---
Action Plan Of GSM Repeater
GSM Repeater 900Mhz