How to annotate SNP variant records in a vcf file with IDs from a bed file?

Question

[William]

How can I annotate a vcf file with IDs from a bed file?

I have a bgzip compressed bed file with the required three columns (chrom, chromStart and chromEnd) and a fourth column that contains the IDs. The bed file does not have a header record.

https://genome.ucsc.edu/FAQ/FAQformat.html

I indexed the file with

tabix -p bed myBedFile.gz

Now I would like to annotate SNP variant records in a vcf file with the IDs from the bed file, based on exact matching of the vcf CHROM / POS fields to the chrom, chromStart and chromEnd columns of the bed file.

I believe this should be possible with bcftools annotate

https://samtools.github.io/bcftools/bcftools.html#annotate

but my

bcftools annotate -c CHROM,FROM,TO,ID -a myBedFile.bed.gz myVcfFile.vcf.gz

returns a vcf file with unchanged SNP variant records, no ID is set in the ID column or the INFO column.

How can I annotate the vcf ID column with IDs from matching positions a bed file? Preferable with a tool like bcftools?

Answer 1

[cpad0112]

I tried running this code (copy/pasted from above except for input files):

bcftools annotate -c CHROM,FROM,TO,ID -a hg19.chr20.refseq.bed.gz na12878.cutadapt.q20.bwa.rmdup.sorted.vcf.gz

Output I got, for a variant with match was:

chr20    62397026    NM_025224    T    C    6.22876e-06    .    AB=0;ABP=0;AC=0;AF=0;AN=2;AO=2;CIGAR=1X;DP=7;DPB=7;DPRA=0;EPP=7.35324;EPPR=13.8677;GTI=0;LEN=1;MEANALT=1;MQM=60;MQMR=60;NS=1;NUMALT=1;ODDS=13.4549;PAIRED=1;PAIREDR=1;PAO=0;PQA=0;PQR=0;PRO=0;QA=9;QR=202;RO=5;RPL=0;RPP=7.35324;RPPR=6.91895;RPR=2;RUN=1;SAF=0;SAP=7.35324;SAR=2;SRF=4;SRP=6.91895;SRR=1;TYPE=snp    GT:DP:RO:QR:AO:QA:GL    0/0:7:5:202:2:9:0,-1.25221,-17.7041

Output I got, for a variant without match was:

chr20    62962891    .    C    T    64.6144    .    AB=0.666667;ABP=4.45795;AC=1;AF=0.5;AN=2;AO=4;CIGAR=1X;DP=6;DPB=6;DPRA=0;EPP=11.6962;EPPR=3.0103;GTI=0;LEN=1;MEANALT=1;MQM=60;MQMR=60;NS=1;NUMALT=1;ODDS=4.49327;PAIRED=1;PAIREDR=1;PAO=0;PQA=0;PQR=0;PRO=0;QA=157;QR=80;RO=2;RPL=0;RPP=11.6962;RPPR=3.0103;RPR=4;RUN=1;SAF=0;SAP=11.6962;SAR=4;SRF=2;SRP=7.35324;SRR=0;TYPE=snp    GT:DP:RO:QR:AO:QA:GL    0/1:6:2:80:4:157:-12.6991,0,-5.78561

I do not see any problem with your code. Check if

• chromosome numbers are same (chrom in VCF and 1st column in bed file).
• VCF and Bed files are indexed with tabix
• bed file corresponds with same genome build and genomic coordinates that span VCF
• indeed there are no matches manually (can happen some times)

[William]

Ok good to known that this should work / works for someone else. I might very well have made a mistake my self. I'll double check what I did and the points you mentioned. Thanks.