Viewing VCF files in IGV

631 views
Skip to first unread message

andrew

unread,
Oct 11, 2011, 12:33:11 PM10/11/11
to igv-help
Hello,
I am attempting to open a VCF file that I have generated from
complete genomics data in IGV. I ran "cgatools testvariants" for
three individuals. I then used testvariants2VCF-v2.pl to create a VCF
file from the output of testvariants. From there, I used bgzip and
tabix on the VCF file that was created. When I finally attempt to
load the .vcf.gz file into IGV, it gives me the following error:

BUG: alleles include some No Calls and some Calls, an illegal state

Any idea how I can deal with this data? Are there certain calls that
I need to remove from the VCF file in order to be able to load it?

Thanks,
Andrew

Jim Robinson

unread,
Oct 11, 2011, 12:52:40 PM10/11/11
to igv-...@googlegroups.com
Hi Andrew,

I'll forward this to the GATK team, who supply the VCF reader. You
might also post on their "Get Satisfaction" forum, but I will post any
response I get here.

Jim

Jim Robinson

unread,
Oct 11, 2011, 1:52:58 PM10/11/11
to igv-...@googlegroups.com
Hi Andrew,

I took this up with the GATK team who provides the VCF readers.  The problem is in the GATK implementation,  it does not handle records for which one allele is called and the other is marked no-call,   e.g.  ./1.     Could you please post this bug in the GATK forum and we can start a discussion there on what to do about it.  They might also have workarounds to suggest.  The forum is at

http://getsatisfaction.com/gsa

-- Jim

andrew

unread,
Oct 11, 2011, 2:15:39 PM10/11/11
to igv-help
Jim,
Thank you very much for the quick reply. I have posted in the GATK
Get Satisfaction forum at:

http://getsatisfaction.com/gsa/topics/the_gatk_vcf_reader_does_not_handle_alleles_that_include_some_no_calls_and_some_calls

Hopefully we can see this resolved. Let me know if I can be of any
assistance.

Thanks,
Andrew

andrew

unread,
Oct 12, 2011, 10:20:34 AM10/12/11
to igv-help
I am also reposting here what I posted in the GATK Get Satisfaction
forum:

For anyone else interested in this - I have found a temporary
workaround. I am not sure what greater impact on the software it has,
but for now it allows me to load the Complete Genomics VCF files.

1) Comment out in GATK in Genotype.java:
if ( nNoCalls > 0 && nNoCalls != alleles.size() )
throw new IllegalArgumentException("BUG: alleles include
some No Calls and some Calls, an illegal state " + this);

2) Comment out in GATK in MutableGenotype.java:
if ( nNoCalls > 0 && nNoCalls != alleles.size() )
throw new IllegalArgumentException("BUG: alleles include
some No Calls and some Calls, an illegal state " + this);

3) Rebuild the vcf.jar in GATK using the command "ant vcf.jar"

4) Drop the jar into IGV project and rebuild

Hope this is helpful to anyone else who has the same issue.
-Andrew

On Oct 11, 2:15 pm, andrew <ashinoh...@gmail.com> wrote:
> Jim,
>    Thank you very much for the quick reply.  I have posted in the GATK
> Get Satisfaction forum at:
>
> http://getsatisfaction.com/gsa/topics/the_gatk_vcf_reader_does_not_ha...

Jim Robinson

unread,
Oct 12, 2011, 11:21:34 AM10/12/11
to igv-...@googlegroups.com, andrew
Thanks Andrew, that is very helpful.

> I am also reposting here what I posted in the GATK Get Satisfaction
> forum:
>
> For anyone else interested in this - I have found a temporary
> workaround. I am not sure what greater impact on the software it has,
> but for now it allows me to load the Complete Genomics VCF files.
>
> 1) Comment out in GATK in Genotype.java:
> if ( nNoCalls> 0&& nNoCalls != alleles.size() )

> throw new IllegalArgumentException("BUG: alleles include
> some No Calls and some Calls, an illegal state " + this);
>
> 2) Comment out in GATK in MutableGenotype.java:
> if ( nNoCalls> 0&& nNoCalls != alleles.size() )

Sean Davis

unread,
Jun 27, 2012, 6:38:15 PM6/27/12
to igv-...@googlegroups.com
Hi, Jim and others.

I saw a comment in the Get Satisfaction forum that a change was committed to GATK a few months ago to fix this problem.  I think I might still be seeing the same problem using IGV 2.1 (downloaded today).  Can anyone comment as to whether this problem is resolved in IGV or does the problem still exist?  I have not applied this patch to GATK/IGV, but I will if the problem is known to still exist.

Thanks.  And sorry to go the lazy route of asking questions first and patching later.

Sean

Sean Davis

unread,
Jun 27, 2012, 7:02:39 PM6/27/12
to igv-...@googlegroups.com
Sorry for the noise.  I had missed a structural variant in the file.  Things are working as expected.  

Sean
Reply all
Reply to author
Forward
0 new messages