解析 VCF 文件的 INFO 字段

Question

Please help me parse a VCF file. I am pasting a real example.

Input:

1   1014143 rs786201005 C   T   .   .   RS=786201005;RSPOS=1014143;dbSNPBuildID=144;SSR=0;SAO=1;VP=0x050068000605000002110100;GENEINFO=ISG15:9636;WGT=1;VC=SNV;PM;PMC;NSN;REF;ASP;LSD;OM;CLNALLE=1;CLNHGVS=NC_000001.11:g.1014143C>T;CLNSRC=OMIM_Allelic_Variant;CLNORIGIN=1;CLNSRCID=147571.0003;CLNSIG=5;CLNDSDB=MedGen:OMIM:Orphanet;CLNDSDBID=C4015293:616126:ORPHA319563;CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification;CLNREVSTAT=no_criteria;CLNACC=RCV000162196.3
1   1014228 rs1921  G   A,C .   .   RS=1921;RSPOS=1014228;dbSNPBuildID=36;SSR=0;SAO=0;VP=0x050328000a0517053f000100;GENEINFO=ISG15:9636;WGT=1;VC=SNV;PM;PMC;S3D;SLO;NSM;REF;ASP;VLD;G5A;G5;HD;GNO;KGPhase1;KGPhase3;CLNALLE=1;CLNHGVS=NC_000001.11:g.1014228G>A;CLNSRC=.;CLNORIGIN=1;CLNSRCID=.;CLNSIG=2;CLNDSDB=MedGen;CLNDSDBID=CN169374;CLNDBN=not_specified;CLNREVSTAT=single;CLNACC=RCV000455759.1;CAF=0.6611,0.3389,.;COMMON=1
1   1014316 rs672601345 C   CG  .   .   RS=672601345;RSPOS=1014319;dbSNPBuildID=142;SSR=0;SAO=1;VP=0x050068001205000002110200;GENEINFO=ISG15:9636;WGT=1;VC=DIV;PM;PMC;NSF;REF;ASP;LSD;OM;CLNALLE=1;CLNHGVS=NC_000001.11:g.1014319dupG;CLNSRC=OMIM_Allelic_Variant;CLNORIGIN=1;CLNSRCID=147571.0002;CLNSIG=5;CLNDSDB=MedGen:OMIM:Orphanet;CLNDSDBID=C4015293:616126:ORPHA319563;CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification;CLNREVSTAT=no_criteria;CLNACC=RCV000148989.5
1   1014359 rs672601312 G   T   .   .   RS=672601312;RSPOS=1014359;dbSNPBuildID=142;SSR=0;SAO=1;VP=0x050068000605000002110100;GENEINFO=ISG15:9636;WGT=1;VC=SNV;PM;PMC;NSN;REF;ASP;LSD;OM;CLNALLE=1;CLNHGVS=NC_000001.11:g.1014359G>T;CLNSRC=OMIM_Allelic_Variant;CLNORIGIN=1;CLNSRCID=147571.0001;CLNSIG=5;CLNDSDB=MedGen:OMIM:Orphanet;CLNDSDBID=C4015293:616126:ORPHA319563;CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification;CLNREVSTAT=no_criteria;CLNACC=RCV000148988.5
1   1020183 rs539283387 G   C   .   .   RS=539283387;RSPOS=1020183;dbSNPBuildID=142;SSR=0;SAO=0;VP=0x050000000a05040026000100;GENEINFO=AGRN:375790;WGT=1;VC=SNV;NSM;REF;ASP;VLD;KGPhase3;CLNALLE=1;CLNHGVS=NC_000001.11:g.1020183G>C;CLNSRC=.;CLNORIGIN=1;CLNSRCID=.;CLNSIG=3;CLNDSDB=MedGen;CLNDSDBID=CN169374;CLNDBN=not_specified;CLNREVSTAT=single;CLNACC=RCV000424799.1;CAF=0.9904,0.009585;COMMON=1
1   1020216 rs764659938 C   G   .   .   RS=764659938;RSPOS=1020216;dbSNPBuildID=144;SSR=0;SAO=0;VP=0x050000000a05040002000100;GENEINFO=AGRN:375790;WGT=1;VC=SNV;NSM;REF;ASP;VLD;CLNALLE=1;CLNHGVS=NC_000001.11:g.1020216C>G;CLNSRC=.;CLNORIGIN=1;CLNSRCID=.;CLNSIG=0;CLNDSDB=MedGen;CLNDSDBID=CN221809;CLNDBN=cancer;CLNREVSTAT=single;CLNACC=RCV000422793.1

And I need an output:

1014143 rs786201005 C   T   CLNSIG=5    CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification
1014228 rs1921  G   A,C CLNSIG=2    CLNDBN=not_specified
1014316 rs672601345 C   CG  CLNSIG=5    CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification
1014359 rs672601312 G   T   CLNSIG=5    CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification
1020183 rs539283387 G   C   CLNSIG=3    CLNDBN=not_specified
1020216 rs764659938 C   G   CLNSIG=0    CLNDBN=not_provided

That means print column 2,3,4,5 and then parse last column and print just CLNSIG and CLNDBN . Problem is, that those values are not always in the same position.

My try was:

awk -v OFS="\t"'{print $2,$3,$4,$5,$8}' input

...and then I have no clue how to get CLNSIG and CLNDBN .

Thank you for any ideas.

Answer 1

1. Pure bash , works by using bash to parse the remaining variables in $h , with parameter tranformation output:

    while read abcdefgh ; do declare ${h//;/ } printf "%s\\t%-10s\\t%s\\t%s\\t%s\\t%s\\n" $b $c $d $e ${CLNSIG@A} ${CLNDBN@A} done < input

   Output:

    1014143 rs786201005 CT CLNSIG='5' CLNDBN='Immunodeficiency_38_with_basal_ganglia_calcification' 1014228 rs1921 GA,C CLNSIG='2' CLNDBN='not_specified' 1014316 rs672601345 C CG CLNSIG='5' CLNDBN='Immunodeficiency_38_with_basal_ganglia_calcification' 1014359 rs672601312 GT CLNSIG='5' CLNDBN='Immunodeficiency_38_with_basal_ganglia_calcification' 1020183 rs539283387 GC CLNSIG='3' CLNDBN='not_specified' 1020216 rs764659938 CG CLNSIG='0' CLNDBN='cancer'

2. POSIX shell, grep and printf method:

    while read abcdefgh ; do printf "%s\\t%-10s\\t%s\\t%s\\t%s\\t%s\\n" $b $c $d $e \\ $( echo "$h" | grep -o 'CLN\\(SIG\\|DBN\\)=[^;]*' ) ; done < input

   Output:

    1014143 rs786201005 CT CLNSIG=5 CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification 1014228 rs1921 GA,C CLNSIG=2 CLNDBN=not_specified 1014316 rs672601345 C CG CLNSIG=5 CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification 1014359 rs672601312 GT CLNSIG=5 CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification 1020183 rs539283387 GC CLNSIG=3 CLNDBN=not_specified 1020216 rs764659938 CG CLNSIG=0 CLNDBN=cancer

Answer 2

With perl

$ perl -lane 'print join "\t",(@F[1..4], /(?:CLNSIG|CLNDBN)=[^;]+/g)' ip.txt
1014143 rs786201005 C   T   CLNSIG=5    CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification
1014228 rs1921  G   A,C CLNSIG=2    CLNDBN=not_specified
1014316 rs672601345 C   CG  CLNSIG=5    CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification
1014359 rs672601312 G   T   CLNSIG=5    CLNDBN=Immunodeficiency_38_with_basal_ganglia_calcification
1020183 rs539283387 G   C   CLNSIG=3    CLNDBN=not_specified
1020216 rs764659938 C   G   CLNSIG=0    CLNDBN=cancer

• -a option to split input on white-space, saved in @F array
• /(?:CLNSIG|CLNDBN)=[^;]+/g will return the CLNSIG and CLNDBN fields
• @F[1..4] gives fields 2nd to 5th (index starts from 0 )
• See http://perldoc.perl.org/perlrun.html#Command-Switches for details on -lane options

Answer 3

It can be done using awk:

script.awk

BEGIN { OFS="\t" }
      { clnsig = clndbn = ""
        if( match( $8, /CLNSIG=[^;]+/ ) ) {
          clnsig = substr( $8, RSTART, RLENGTH )
        }
        if( match( $8, /CLNDBN=[^;]+/ ) ) {
          clndbn = substr( $8, RSTART, RLENGTH )
        }
        print $2, $3, $4, $5, clnsig, clndbn
      }

Or more compact, in case that CLNDBN is always after CLNSIG :

script.awk

BEGIN { OFS="\t" }
    { match($8,/(CLNSIG=[^;]+).*(CLNDBN=[^;]+)/, tmp) 
      print $2,$3,$4,$5, tmp[1], tmp[2]
    }

The function match matches a regular expression. The first form sets the variables RSTART and RLENGTH so that you can extract the text with substring .

The second form puts the first subexpression (first parentheses) in the array tmp at pos 1, the second subexpression at pos 2 and so on.

The regular expression CLNSIG=[^;]+ matches a literal CLNSIG= followed by a substring up to (but not including) the ; .