gatk GenotypeGVCFs USER ERROR: The list of input alleles must contain <NON_REF> as an allele but that is not the case

[lubocoix]

I have already get ".g.vcf " through "gatk Haplotype Caller" but when I used the code "./gatk GenotypeGVCFs -R /Users/lubo/sorgum/GCF_000003195.3_Sorghum_bicolor_NCBIv3_genomic.fna.
-V /Users/lubo/sorgum/propinquum_variation.g.vcf
-O /Users/lubo/sorgum/propinquum.vcf" to generate the output file "propinquum.vcf" ,A USER ERROR has occurred: The list of input alleles must contain <NON_REF> as an allele but that is not the case at position 11733; please use the Haplotype Caller with gVCF output to generate appropriate records。
I don't know what's wrong with my code ,is that mean my input file have something wrong?

[droazen]

@lubocoix Are you running HaplotypeCaller with the -ERC GVCF argument?

[ShuwenXia]

Hi lubocoix, did you fix the error? I got the same issue. Please let me know how did you fix the problem. Many Thanks!!

[lubocoix]

I have solved my problem, I need see your code that I can know what’s wrong with it.

…

2021年10月8日 上午11:13，ShuwenXia ***@***.***> 写道： Hi lubocoix, did you fix the error? I got the same issue. Please let me know how did you fix the problem. Many Thanks!! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#7147 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ATGHJIZQKEUD7NC3KMLVOZ3UFZOUHANCNFSM4ZICDOSA>. Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

[ShuwenXia]

I have used Haplotype Caller to get husheep_reseq.g.vcf file and run "./share/nas1/comput5/Tools/GATK/gatk-4.2.0.0/gatk --java-options -Xmx80G GenotypeGVCFs -R /share/nas1/comput5/ref_genome/husheep_ref/GCA_011170295.1_ASM1117029v1_genomic.fna -V husheep_reseq.g.vcf -O husheep_reseq.vcf “ to generate husheep_reseq.vcf file, but I got an error "The list of input alleles must contain <NON_REF> as an allele but that is not the case at position 100; please use the Haplotype Caller with gVCF output to generate appropriate records". Seems similar in your case.

[PlatonB]

@droazen In my case, CombineGVCFs crashes on multiallelic variants.

#CHROM  POS     ID      REF     ALT     QUAL    FILTER  INFO    FORMAT  mother_DV
chrX    10001   .       C       <NON_REF>       0       .       END=10060       GT:GQ:MIN_DP:PL 0/0:1:0:0,0,0
chrX    10061   .       G       <NON_REF>       0       .       END=10068       GT:GQ:MIN_DP:PL 0/0:6:2:0,6,59
chrX    10069   .       T       A,<NON_REF>     14.7    PASS    .       GT:GQ:DP:AD:VAF:PL      1/1:13:2:0,2,0:1,0:14,18,0,990,990,990
chrX    10070   .       G       <NON_REF>       0       .       END=10070       GT:GQ:MIN_DP:PL 0/0:6:2:0,6,59
chrX    10071   .       G       <NON_REF>       0       .       END=10111       GT:GQ:MIN_DP:PL 0/0:15:5:0,15,149
chrX    10112   .       G       <NON_REF>       0       .       END=10120       GT:GQ:MIN_DP:PL 0/0:18:6:0,18,179
<...>

org.broadinstitute.hellbender.exceptions.GATKException: Exception thrown at chrX:10071 [VC /home/pbykadorov/family/02_merged/mother_DV_chrX.g.vcf.gz @ chrX:10071-10111 Q0.00 of type=SYMBOLIC alleles=[G*, <NON_REF>] attr={END=10111} GT=GT
:GQ:MIN_DP:PL   0/0:15:5:0,15,149 filters=

Caused by: org.broadinstitute.hellbender.exceptions.UserException: The list of input alleles must contain <NON_REF> as an allele but that is not the case at position 10069; please use the Haplotype Caller with gVCF output to generate app
ropriate records

Note.
<NON_REF> was inserted in place of DeepVariant's <*> manually:

sed -e "s/<\*>/<NON_REF>/g"

[gokalpcelik]

Hi @PlatonB
GATK GVCF tools are designed to work with only GVCFs from GATK HaplotypeCaller. If you wish to work with GVCFs generated by deepvariant or other tools you may use glnexus.

[PlatonB]

@gokalpcelik The same error occurs when working with HaplotypeCaller results.

[gokalpcelik]

Can you also post a line from HaplotypeCaller generated GVCF that causes this issue ?

[PlatonB]

chrX 10069 . T A,<NON_REF> 63.31 . DP=3;ExcessHet=0;MLEAC=1,0;MLEAF=0.5,0;RAW_MQandDP=5337,3 GT:<...>

[gokalpcelik]

How was this gvcf created can you post us your steps as well? Also are you trying to combine gvcfs generated by different tools with haplotypecallers gvcf ?
What version of GATK are you using?

[PlatonB]

How was this gvcf created can you post us your steps as well?

I'm afraid that's non-public information. Perhaps after consulting with my employer, I will publish some of the pipelines' code.

Also are you trying to combine gvcfs generated by different tools with haplotypecallers gvcf ?

Yes, I am merging the gVCFs generated by HaplotypeCaller, DeepVariant, BCFtools mpileup&call and FreeBayes. Before merging, I try to unify the source gVCFs using self-written Bash and Python scripts.

What version of GATK are you using?

gatk --version

Using GATK jar /home/pbykadorov/miniconda3/share/gatk4-4.4.0.0-0/gatk-package-4.4.0.0-local.jar
Running:
    java -Dsamjdk.use_async_io_read_samtools=false -Dsamjdk.use_async_io_write_samtools=true -Dsamjdk.use_async_io_write_tribble=false -Dsamjdk.compression_level=2 -jar /home/pbykadorov/miniconda3/share/gatk4-4.4.0.0-0/gatk-package-4.4.0.0-local.jar --version
The Genome Analysis Toolkit (GATK) v4.4.0.0
HTSJDK Version: 3.0.5
Picard Version: 3.0.0

[gokalpcelik]

This is the keypoint.
"Merging the gVCFs generated by HaplotypeCaller, DeepVariant, BCFtools mpileup&call and FreeBayes. "

CombineGVCFs tool's aim is not to combine GVCFs from different sources. The tool can only work with GVCFs from HaplotypeCaller.

For us to understand the issue clearly we need to get the actual command line for HaplotypeCaller for generating the GVCF file. GVCFs generated by other tools are not our concern therefore if you are doing something outside of GATK domain we cannot help.

There may be other tools doing this kind of work but we don't have extensive information on them and we cannot debug their issues either.

[lbergelson]

@PlatonB Since this is a complex issue involving multiple non-gatk tools and a proprietary pipeline we can't really do much to debug it without a way to reproduce the problem. If you could provide (ideally minimal) inputs to reproduce the issue that would give us much more of a handle to look into it. Otherwise I don't think we can do much with the information you provided.