Numeric genotype to letters

I have an input file like this:

#CHROM  START   ID  Ref Alt 108 139 159 265 350
5   5571    snp_5_5571  C   T   0   0   0   1   0
3   11641   snp_3_11641 T   G   0   1   1   2   2
3   14240   snp_3_14240 G   A,T 0   0   0   0   1

From column 5 onwards if its 0 replace it with ref+ref, if its one replaces it with ref+alt and if its 2 replace it with alt+alt so that the above table should be like that:

#CHROM  START   ID  Ref Alt 108 139 159 265 350
5   5571    snp_5_5571  C   T   CC  CC  CC  CT  CC
3   11641   snp_3_11641 T   G   TT  TG  TG  GG  GG
3   14240   snp_3_14240 G   A   GG  GG  GG  GG  GA

snp

• 1,845 views • link

updated 7.1 years ago by JC

• written 7.1 years ago by waqaskhokhar999

3 answers

 awk '/^#/{print;next;} {split($5,a,/[,]/);for(i=1;i<6;i++) printf("%s%s",i==1?"":"\t",$i);for(i=6;i<=NF;i++) { printf("\t"); if($i==0) printf("%s%s",$4,$4); else if($i==1) printf("%s%s",$4,a[1]); else if($i==2) printf("%s%s",a[1],a[1]); else printf("??");} printf("\n");} ' input.tsv

#CHROM START ID Ref Alt 108 139 159 265 350
5   5571    snp_5_5571  C   T   CC  CC  CC  CT  CC
3   11641   snp_3_11641 T   G   TT  TG  TG  GG  GG
3   14240   snp_3_14240 G   A,T GG  GG  GG  GG  GA

• 0 views • link

written 7.1 years ago by Pierre Lindenbaum

I have no idea what the format of your object is but here is a potential solution in R.

Assuming you have your input file in R and it looks as follows (you can use read.delim()):

  X.CHROM START          ID Ref Alt X108 X139 X159 X265 X350
1       5  5571  snp_5_5571   C   T    0    0    0    1    0
2       3 11641 snp_3_11641   T   G    0    1    1    2    2
3       3 14240 snp_3_14240   G A,T    0    0    0    0    1

Then you can use the following function on this object (class data.frame) in the following way:

convertSNP <- function(var.row){
  ref <- var.row['Ref']
  alt <- var.row['Alt']
  ## keep only first variant
  alt <- gsub(",[AGTC]+","",alt)
  var.row[var.row == 0] <- paste0(ref,ref)
  var.row[var.row == 1] <- paste0(ref,alt)
  var.row[var.row == 2] <- paste0(alt,alt)
  ## get rid of whitespace
  var.row <- gsub(" ","",var.row)
  return(var.row)
}

## run function on data.frame by row and transpose the result
t(apply(df,1,convertSNP))

Giving the result:

     X.CHROM START   ID            Ref Alt   X108 X139 X159 X265 X350
[1,] "5"     "5571"  "snp_5_5571"  "C" "T"   "CC" "CC" "CC" "CT" "CC"
[2,] "3"     "11641" "snp_3_11641" "T" "G"   "TT" "TG" "TG" "GG" "GG"
[3,] "3"     "14240" "snp_3_14240" "G" "A,T" "GG" "GG" "GG" "GG" "GA"

• 0 views • link

written 7.1 years ago by benformatics

$ perl -lae 'for ($i=5;$i<=$#F;$i++) { $F[$i]=$F[3]x2 if ($F[$i]==0); $F[$i]=$F[3].$F[4] if ($F[$i]==1); $F[$i]=$F[4]x2 if ($F[$i]==2); $F[$i]=~ s/,.+//; } print join "\t", @F' < in
#CHROM  START   ID      Ref     Alt     108     139     159     265     350
5       5571    snp_5_5571      C       T       CC      CC      CC      CT      CC
3       11641   snp_3_11641     T       G       TT      TG      TG      GG      GG
3       14240   snp_3_14240     G       A,T     GG      GG      GG      GG      GA

• 0 views • link

written 7.1 years ago by JC