vcfeval Error: No sample name provided but calls is a multi-sample VCF

Question

[simo017]

Hi, I am trying to compare a group of vcf files alltegether against the GIAB vcf file, whene I compared them against the GIAB vcf file one by one everything went fine but when I tried to merge them and then compare the merged file agianst the GIAB vcf I got an error the say: No sample name provided but calls is a multi-sample VCF.

I found a similar post here but with no clue:

Here is my commands: to merge them:

 ./rtg vcfmerge -o merged_extra.vcf.gz full_variant_table_3.vcf.gz full_variant_table_4.vcf.gz full_variant_table_5.vcf.gz full_variant_table_6.vcf.gz full_variant_table_7.vcf.gz full_variant_table_8.vcf.gz

for comparaison:

./rtg vcfeval -t /home/variants/1000g_v37_phase2.sdf -b /home/rtg-tools-3.8.4/GIAB_MY_region.vcf.gz -c /home/variants/vcf_files/merged_extra.vcf.gz -o /home/variants/vcf_result

this gives: Error: No sample name provided but calls is a multi-sample VCF.

I tried to use --sample=<calls_samplename> parameters however I am still getting the same error:

./rtg vcfeval -t /home/variants/1000g_v37_phase2.sdf -b /home/rtg-tools-3.8.4/GIAB_MY_region.vcf.gz -c /home/variants/vcf_files/merged_extra.vcf.gz -o /home/variants/vcf_result --sample GIAB_MY_region,merged_extra

Any help is very appreciated

Answer 1

[geocarvalho]

Your VCF header label has the same name of the file? For example: #CHROM POS ID REF ALT QUAL FILTER INFO FORMAT GIAB_MY_region #CHROM POS ID REF ALT QUAL FILTER INFO FORMAT merged_extra

or is another? Because the label expected for the --sample argument is the label present in the VCF header.

best regards.

Answer 2

[simo017]

Hi geocarvalho, Thank you for your answer.

[geocarvalho]

About your other comment. If you have a VCF with mulitple samples you must choose one at a time. For example:

• For a header: #CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_1 Sample_2 Sample_3

command: ./rtg vcfeval --baseline=reference.vcf.gz --bed-regions=roi.bed -c merged_extra.vcf.gz -o multi_vcf -t /path/to/genome.fasta.sdf --sample=Reference_label,Sample_2

• But if you have a multi VCF with the same sample you should rename your initial VCFs to the same label. For example:

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_1

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_2

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_3

Should be:

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_1

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_1

#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT Sample_1

To re-run the rtg vcfmerge

• Or you can simple use GATK CombineVariants with the genotypeMergeOptions PRIORITIZE parameter (it's what I recommend) and isn't necessary change the VCF's header sample label.

• And then you can run rtg vcfeval ./rtg vcfeval --baseline=reference.vcf.gz --bed-regions=roi.bed -c merged_extra.vcf.gz -o multi_vcf -t /path/to/genome.fasta.sdf --sample=Reference_label,Sample_1