awk or perl to update/replace pattern with specific value - perl

Question

Trying to use perl or awk to update a specific pattern in a file to a specific text value. The code does execute but does not produce the desired output, rather it seems to repeat $3 instead of update it. Thank you :).
So when 0/0 or 1/1 is found in $3 it is updated/replaced with hom or when 0/1 is found in $3 it is replaced/updated to het.
file space delimited
chr1 115252242 0/0
chr1 247587408 1/1
chr5 35873605 0/1
perl
perl -plae '
BEGIN{ %h = qw(0/0 hom 0/1 het 1/1 hom 1/2 het 2/2 hom) } # define patterns
/^.*([0-2]\/[0-2])/ # define hash values to use with patterns and
$_ .= join "\t", ("", $1, $2, $h{$3})' file # print updated output
current
chr1 115252242 0/0 0/0
chr1 247587408 1/1 1/1
chr5 35873605 0/1 0/1
desired tab-delimited
chr1 115252242 hom
chr1 247587408 hom
chr5 35873605 het

Answer 1

Your Perl script seems to be trying to use $1, $2 as if they are awk fields:
/^.*([0-2]\/[0-2])/ and
$_ .= join "\t", ("", $1, $2, $h{$3})' file
However, in Perl, these variables refer to capture groups from a regex match. In your regex, there is only a single capture group. This is clearer, if you modify the join as:
/^.*([0-2]\/[0-2])/ and
$_ .= join "#", ("", ":", $1, "::", $2, ":::", $h{$3})' file
Your output will be:
chr1 115252242 0/0#:#0/0#::##:::#
chr1 247587408 1/1#:#1/1#::##:::#
chr5 35873605 0/1#:#0/1#::##:::#
In addition .= appends to the existing value; it does not replace it.
A modification to your script to fix this is to add an extra capture group:
/^(.*)([0-2]\/[0-2])/
Now you can refer to $1 and $2:
$_ = join("\t", $1, $h{$2});
However, as $1 ends with whitespace already, and you seem to want to end up with TSV, you can extract leading/trailing whitespace by:
/^\s*(.*)([0-2]\/[0-2])\s*$/
and then replace all remaining runs of whitespace with a single tab:
s/\s+/\t/g
Finally, instead of m// (match) followed by join, you can simply use s/// (find and replace). We use && so that the second replacement only happens if the first one did anything:
s/^\s*(.*)([0-2]\/[0-2])\s*$/$1$h{$2}/ && s/\s+/\t/g
The final program is:
perl -plae '
BEGIN{ %h = qw( 0/0 hom 0/1 het 1/1 hom 1/2 het 2/2 hom ) }
s/^\s*(.*)([0-2]\/[0-2])\s*$/$1$h{$2}/ && s/\s+/\t/g
' file
An awk equivalent might be:
awk -v OFS="\t" '
$3 ~ /^[0-9]+\/[0-9]+/$/ {
split($3,n,"/")
$3 = n[1]==n[2] ? "hom" : "het"
}
1
' file
This processes lines with whitespace-delimited columns. It checks if the 3rd column has the appropriate form. If so, it is split into two numbers. If the numbers match, it sets the 3rd column to "hom"; if not, to "het". 1 is a shorter way to write {print}.

Answer 2

Could you please try following.
awk '
BEGIN{
OFS="\t"
}
$NF=="0/0" || $NF=="1/1"{
$NF="hom"
}
$NF=="0/1"{
$NF="het"
}
1
' Input_file
Explanation: Adding explanation for above code.
awk ' ##Starting awk program here.
BEGIN{ ##Starting BEGIN section of this program here.
OFS="\t" ##Setting OFS as \t(tab) here.
} ##Closing BLOCK for BEGIN section of this command here.
$NF=="0/0" || $NF=="1/1"{ ##Checking condition if last field is either equal to 0/0 OR 1/1 then do following.
$NF="hom" ##Set last field value of as string hom string here.
} ##Closing BLOCK for this above condition here.
$NF=="0/1"{ ##Checking condition if last field value is 0/1 then do following.
$NF="het" ##Setting last field value as het string here.
} ##Closing BLOCK for condition here.
1 ##mentioning 1 will print edited/non-edited line here.
' Input_file ##Mentioning Input_file name here.

Answer 3

Following code should give desired result
use strict;
use warnings;
my $file = do { local $/; <DATA> };
$file =~ s#(0/0|1/1)#hom#g;
$file =~ s#0/1#het#g;
$file =~ s# +#\t#g; # replace space separators to tab
print $file;
__DATA__
chr1 115252242 0/0
chr1 247587408 1/1
chr5 35873605 0/1
output
chr1 115252242 hom
chr1 247587408 hom
chr5 35873605 het

Answer 4

$ awk 'BEGIN{map["0/0"]=map["1/1"]="hom"; map["0/1"]="net"} $3 in map{$3=map[$3]} 1' file
chr1 115252242 hom
chr1 247587408 hom
chr5 35873605 net