max size for a Tcl value (2147483647 bytes) exceeded

[ekarlins]

I'm running v2.2 of AnnotSV on a SGE cluster. I'm running on a couple of gene lists separately using "-candidateGenesFile" and "-candidateGenesFiltering". The run completes for a small gene list, but fails for the larger gene list and exits abruptly. The error message is in the subject line. I tried giving more memory to the job (200G), but that doesn't fix it as there doesn't seem to be a way to specify to AnnotSV (or Tcl) to use more memory. I'm going to try splitting my gene list into two and see if that works. Is there another fix for this?

Thanks,
Eric

[lgmgeo]

Hi Eric,

Can you tell me the size of your candidateGenesFile?
And how many lines are there ?

Best regards,
Véronique

[ekarlins]

Véronique,
Thanks for your quick reply! The original candidateGenesFile has 1730 lines, but I've since split it into smaller parts of 346 lines and it still fails with the same error.
Is there any way to tell Tcl to use more memory?

Thanks,
Eric

[lgmgeo]

Ok, one more question:
Are the gene names space-separated, tabulation-separated, or line-break-separated?

[lgmgeo]

Can you contact me by email: [email protected] to continue the debbuging?

[ekarlins]

Ok. I sent you an email. Thanks!

[lgmgeo]

Conclusion on this post:

• Running AnnotSV without using the candidateGenes flags fails as well with the same error.
• AnnotSV was used to annotate a very large SV input file.

To solve this, you can split your input file into smaller files, run AnnotSV and then later merge them into a single output file.

This will be fixed in a future release.
Sorry for any inconvenience that may cause.

[Mkddb]

Hello Veronique,

Hope you are doing great.
I also had a similar issue a while ago, so posting my query in the same thread.

You are right. AnnotSV doesn't annotate a very large SV input file, say a merged file from multiple genomes or a large scale study. can we expect a better resolution around this in any upcoming version soon.
or shall i continue with annotation of split input file and later merging.

Thanks in Advance

[lgmgeo]

Hi Mkddb,

Good news, a new collaboration is just being set up to implement AnnotSV in Python. It should solve this bug!
But it may take a while, I can't say exactly how long.

I will keep this thread open until then.

Best,
Véronique

[Mkddb]

Okay Great.

That's a good news. Good luck with the new collaboration.
Thanks for informing.

[Stikus]

Hello @lgmgeo
We experienced the same problem:

AnnotSV 3.0

Copyright (C) 2017-2020 GEOFFROY Veronique

Please feel free to contact me for any suggestions or bug reports
email: [email protected]

Tcl/Tk version: 8.6

Application name used (defined with the "ANNOTSV" environment variable):
/soft/AnnotSV-3.0


...downloading the configuration data (February 21 2021 - 20:49)
        ...configuration data by default
        ...configuration data from /soft/AnnotSV-3.0/etc/AnnotSV/configfile
        ...configuration data given in arguments
        ...checking all these configuration data

...checking the annotation data sources

...listing arguments
        ******************************************
        AnnotSV has been run with these arguments:
        ******************************************
        -REreport no
        -SVinputFile /var/lib/cwl/stg3e23f422-fa51-450a-906d-1d842ff98155/77050025_NEB_manta.candidateSV.vcf
        -SVinputInfo 1
        -SVminSize 50
        -annotationMode both
        -annotationsDir /var/lib/cwl/stgff43072e-534a-4e34-a491-3fde1a061db3/ref/AnnotSV/3.0
        -bcftools bcftools
        -bedtools bedtools
        -candidateGenesFiltering no
        -genomeBuild GRCh38
        -includeCI yes
        -metrics us
        -minTotalNumber 500
        -organism Human
        -outputDir /LlySzz/output
        -outputFile 77050025_NEB.annotsv.tsv
        -overlap 100
        -overwrite yes
        -promoterSize 500
        -rankFiltering 1 2 3 4 5
        -reciprocal no
        -samplesidBEDcol -1
        -snvIndelPASS 0
        -svtBEDcol -1
        -tx RefSeq
        ******************************************

max size for a Tcl value (2147483647 bytes) exceeded

Will wait for Python implementation and fix.

[nvnieuwk]

Hi, any news on the python version? We're also hitting this maximum value error. I'll implement a temporary splitting solution for now, but that isn't really the best for SV VCFs with a lot of BNDs 😁

[lgmgeo]

but that isn't really the best for SV VCFs with a lot of BNDs

You're right.

Ok, definitely, I have to admit that I can't find the time to reimplement AnnotSV in Python.

I'll look at this bug from another angle and see how to fix it in Tcl code. Not in the near future but as soon as possible

[nvnieuwk]

Thanks for your response!
Don't worry, it's not that urgent since I've implemented a workaround for now in my pipeline. (Splitting per SVTYPE)