extract and generate a VCF file with only "PASS"

Question

[zengtony743]

I have a VCF file and I want to generate a new VCF file with the variants which have only FILTER as "PASS" left

I tried

 1) grep 4751snpf.vcf > 4751PASS.vcf | grep “FILTER=PASS”
 2) grep 4751snpf.vcf > 4751PASS.vcf | grep “PASS”

They both failed. Here is the VCF file looks like:

#CHROM  POS ID  REF ALT QUAL    FILTER  INFO    FORMAT  MO_4751

chr10   3124905 .   A   C   7797.30 PASS    ABHet=0.588;ABHom=0.995;
AC=1;AF=0.500;AN=2;BaseQRankSum=-4.919;DP=8;Dels=0.00;FS=7.563;HaplotypeScore=0.
4537;InbreedingCoeff=-0.5755;MQ=55.56;MQ0=1;MQRankSum=-4.937;OND=3.265e-03;QD=12
.97;ReadPosRankSum=5.683    GT:AD:DP:GQ:PL  0/1:7,1:8:9:9,0,228
chr10   3176529 .   G   T   6361.10 PASS    ABHet=0.651;ABHom=0.864;
AC=1;AF=0.500;AN=2;BaseQRankSum=-9.379;DP=14;Dels=0.01;FS=0.292;HaplotypeScore=5
.7125;InbreedingCoeff=-0.7938;MQ=55.81;MQ0=0;MQRankSum=-7.360;OND=0.035;QD=8.60;
ReadPosRankSum=-2.204   GT:AD:DP:GQ:PL  0/1:9,5:14:99:109,0,266
chr10   3257558 .   A   C   349.31  LowCoverage ABHet=0.543;ABHo
m=1.00;AC=1;AF=0.500;AN=2;BaseQRankSum=-1.917;DP=3;Dels=0.00;FS=1.412;HaplotypeS
core=0.4039;InbreedingCoeff=0.0676;MQ=54.44;MQ0=0;MQRankSum=-2.510;QD=15.88;Read
PosRankSum=-1.127   GT:AD:DP:GQ:PL  0/1:2,1:3:26:26,0,63
chr10   3505793 .   A   C   406.07  LowCoverage ABHet=0.556;ABHo
m=0.885;AC=1;AF=0.500;AN=2;BaseQRankSum=-0.055;DP=2;Dels=0.00;FS=136.819;Haploty
peScore=0.2675;InbreedingCoeff=-0.1981;MQ=48.36;MQ0=0;MQRankSum=-3.192;OND=0.090
;QD=7.66;ReadPosRankSum=-1.637  GT:AD:DP:GQ:PL  0/1:1,1:2:21:21,0,27

[zengtony743]

Just found that vcftool has --remove-filtered-all. Has anyone tried it?

Answer 1

[finswimmer]

$ bcftools view -i 'ID="PASS"' input.vcf > output.vcf

[Ahmed Alhendi]

I like bcftools, it is so fast and can handel larege vcf files from whole genome sequencing (WGS) with high efficiency. However, the bcftools option -i 'ID="PASS"' will not work for all versions of vcf files, at least it did not work with me with illumenia WGS SNV.vcf files. To make sure it will work with all types of vcf files, I would use the following command instead:

bcftools view -f PASS input.vcf > output.vcf -Ahmed Alhendi

Answer 2

[Vivek]

You are grepping the file name instead of the pattern and writing it to another file before using a pipe to grep for the pattern you want. Try this command instead.

awk -F '\t' '{if($0 ~ /\#/) print; else if($7 == "PASS") print}' 4751snpf.vcf > 4751PASS.vcf

Edited to add an apastrophe.

[zengtony743]

Thanks, Vivek. I am running the one that you posted It shows this, Is it normal?

$ awk -F '\t' '{if($0 ~ /#/) print; else if($7 == "PASS") print} 4751snpf.vcf > 4751PASS1.vcf &

[spvensko]

There's an apostrophe missing between "print}" and "4751". I think the correct syntax is:

awk -F '\t' '{if($0 ~ /\#/) print; else if($7 == "PASS") print}' 4751snpf.vcf > 4751PASS1.vcf

[Vivek]

Ah thank you, I typed that post right before leaving home and didn't get a chance to check the error.

Answer 3

[spvensko]

I believe this should work:

awk '$7=="PASS" {print $0}' 4751snpf.vcf > 4751PASS.vcf

[zengtony743]

Thanks, spvensko, it works by producing the following file, however, all information on the header line lost too

chr10   3176529 .   G   T   6361.10 PASS    ABHet=0.651;ABHom=0.864;
AC=1;AF=0.500;AN=2;BaseQRankSum=-9.379;DP=14;Dels=0.01;FS=0.292;HaplotypeScore=5
.7125;InbreedingCoeff=-0.7938;MQ=55.81;MQ0=0;MQRankSum=-7.360;OND=0.035;QD=8.60;
ReadPosRankSum=-2.204   GT:AD:DP:GQ:PL  0/1:9,5:14:99:109,0,266
chr10   3696700 .   T   G   988.21  PASS    ABHet=0.604;ABHom=0.869;
AC=1;AF=0.500;AN=2;BaseQRankSum=1.305;DP=5;Dels=0.00;FS=387.692;HaplotypeScore=0
.3825;InbreedingCoeff=-0.3615;MQ=49.17;MQ0=0;MQRankSum=-5.599;OND=0.066;QD=5.95;
ReadPosRankSum=-7.691   GT:AD:DP:GQ:PL  0/1:3,2:5:44:44,0,61
chr10   4033401 .   G   A,T 16302.20    PASS    AC=1,0;AF=0.500,
0.00;AN=2;BaseQRankSum=-7.317;DP=34;Dels=0.00;FS=235.717;HaplotypeScore=12.5074;
InbreedingCoeff=-0.4964;MQ=55.46;MQ0=1;MQRankSum=-29.193;QD=6.18;ReadPosRankSum=
-5.443  GT:AD:DP:GQ:PL  0/1:27,7,0:34:99:111,0,818,186,836,1023
chr10   4403027 .   A   G   3098.16 PASS    ABHet=0.673;ABHom=0.977;
AC=1;AF=0.500;AN=2;BaseQRankSum=6.290;DP=10;Dels=0.00;FS=0.000;HaplotypeScore=6.
4682;InbreedingCoeff=-0.4427;MQ=49.74;MQ0=0;MQRankSum=5.366;OND=8.775e-03;QD=7.9
4;ReadPosRankSum=-9.160 GT:AD:DP:GQ:PL  0/1:8,2:10:42:42,0,149

[spvensko]

You should follow Vivek's approach (A: extract and generate a VCF file with only "PASS") as it maintains the header lines.

[zengtony743]

Thanks, Vivek. I am running the one that you posted It shows this, Is it normal?

$ awk -F '\t' '{if($0 ~ /\#/) print; else if($7 == "PASS") print} 4751snpf.vcf > 4751PASS1.vcf &
> 
> 
> 
> 
> 
> 
> 
> 
>

Answer 4

[Lakshman Teja]

Hello,
You can try the below GATK command to filter variants by 'PASS':

gatk --java-options '-Xmx20G -XX:+UseParallelGC -XX:ParallelGCThreads=8' SelectVariants -R reference.fasta -V snps.vcf.gz --exclude-filtered true -O filter.vcf.gz

Alternatively, you can use vcffilter tools:

vcffilter -f "FILTER = PASS" snps.vcf.gz > filter.vcf.gz

Hope this works

-Lakshman

Answer 5

[Ahmed Alhendi]

If you only want to filter for PASS records then you don't need to use 'AWK '. This simply can be done using bcftools.

bcftools view -f PASS input.vcf > output.vcf

Answer 6

[osowiecki]

$ zcat ./4_strelka.vcf.gz | grep -P "\tPASS\t|^#" > test

Answer 7

[carmen.bugarin.diz]

vcftools --remove-filtered-all --recode 4751snpf.vcf > 4751PASS.vcf