Assertion failed in readjoiner assembly step

[hp2048]

Hi, While running the gt readjoiner assembly, I received the following error.

Assertion failed: (nofreads == nofreads_i), function gt_readjoiner_assembly_build_contained_reads_list, file src/tools/gt_readjoiner_assembly.c, line 472.`

gt version info:

gt (GenomeTools) 1.5.8
Copyright (c) 2003-2016 G. Gremme, S. Steinbiss, S. Kurtz, and CONTRIBUTORS
Copyright (c) 2003-2016 Center for Bioinformatics, University of Hamburg
See LICENSE file or http://genometools.org/license.html for license details.
Used compiler: cc (Ubuntu 4.9.1-16ubuntu6) 4.9.1
Compile flags:  -g -Wall -Wunused-parameter -pipe -fPIC -Wpointer-arith -O3 -m32 -m64 -Werror

I am running the command on CentOS release 6.6
Pre filter, overlap and assembly commands were as follows:

gt readjoiner prefilter -v -readset readset1 -memdes -maxlow 10 -lowqual 10 -db Readset1.fastq Readset2.fastq Readset3_1.fastq:Readset3_2.fastq:200
gt -j 12 readjoiner overlap -readset readset -l 35
gt readjoiner assembly -readset algae4639 -l 35 -errors -spmfiles 12 -v

Could you please help me solve this problem?
Looking forward to hearing from you.
Kind regards
hardip

[satta]

Hi @patelhardip -- I have looked at this piece of code and it looks like that an actual error occurring during a previous operation (see https://github.com/genometools/genometools/blob/master/src/tools/gt_readjoiner_assembly.c#L472) may have been obscured by this assertion. If you are comfortable with using Git, could you perhaps try pulling from the 'readjoiner_test_fix' branch from my own repository (https://github.com/satta/genometools/tree/readjoiner_test_fix), build the code and retry your run with this version?
@ggonnella -- would you be OK with this? It looks like the call to gt_cntlist_parse() in line 470 can fail and this is not addressed in the assertion. It's just a hunch, but the simplest explanation. See satta@664f245.

[hp2048]

Dear Sascha
Thank you so much for the response. I am currently away from office but will give this a try as soon as I can and report back.
kind regards

Hardip Patel

On 4 Jul 2016, at 7:12 PM, Sascha Steinbiss notifications@github.com wrote:

Hi @patelhardip https://github.com/patelhardip -- I have looked at this piece of code and it looks like that an actual error occurring during a previous operation (see https://github.com/genometools/genometools/blob/master/src/tools/gt_readjoiner_assembly.c#L472 https://github.com/genometools/genometools/blob/master/src/tools/gt_readjoiner_assembly.c#L472) may have been obscured by this assertion. If you are comfortable with using Git, could you perhaps try pulling from the 'readjoiner_test_fix' branch from my own repository (https://github.com/satta/genometools/tree/readjoiner_test_fix https://github.com/satta/genometools/tree/readjoiner_test_fix), build the code and retry your run with this version?
@ggonnella https://github.com/ggonnella -- would you be OK with this? It looks like the call to gt_cntlist_parse() in line 470 can fail and this is not addressed in the assertion. It's just a hunch, but the simplest explanation.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub #806 (comment), or mute the thread https://github.com/notifications/unsubscribe/AJqmvPkBfzlJJME7ROCAR9B6PxWjOhnRks5qSM6SgaJpZM4JCvRj.

[ggonnella]

@satta: Yes, thank you for fixing it.

However, I am not sure, why gt_cntlist_parse() fails, in this case.

@patelhardip: Maybe we know more, when you have tried the fixed version. Thank you for reporting the problem.

[ggonnella]

@patelhardip:

Were the commands really exactly these?

gt readjoiner prefilter -v -readset readset1 ...
gt -j 12 readjoiner overlap -readset readset ...
gt readjoiner assembly -readset algae4639 ...

The readset name must be the same in all steps, otherwise it will fail, as the necessary files are not found! This could explain the problem.

[hp2048]

@ggonnella I have checked my previous commands and I can confirm that readset names are identical in all three commands.
@satta Unfortunately I was not able to pull your readjoiner_test_fix to test it. I believe that the fix is merged with the main repo. So I nuked my local repo and started from scratch.
Following commands were used:
gt readjoiner prefilter -v -readset algaetest -memdes -maxlow 10 -lowqual 10 -db algae_1_4639.fastq algae_2_4639.fastq algae_3_4639_1.fastq:algae_3_4639_2.fastq:200
gt -j 12 readjoiner overlap -readset algaetest -l 35
gt readjoiner assembly -readset algaetest -l 35 -errors -spmfiles 12 -v

This time it failed with the following error:
gt readjoiner assembly: error: fopen(): cannot open file 'algaetest.1.cnt': No such file or directory
Assuming the error was stemming from absence of files, I removed the -j12 and -spmfiles 12 options from overlap and assembly tools respectively.

I was able to successfully run the program. I guess the problem lies in the data. i.e. I don't have sufficient reads to partition into 12 bins and therefore .cnt files were not created.

Anyways, I will trial out without the threaded option and see if it works out with my remaining data and update you.

Hardip

[satta]

@patelhardip I also suspect that it is related to your data. Please note that the -j12 should only make a difference if you compiled with make ... threads=yes which I'm not sure you did. I guess I can't say much about -spmfiles, @ggonnella probably knows more.

However, at least you are getting the correct error message now, which I already consider a bug fixed ;)