My suggestion is to modify the things on output instead of modifying the things in your data structure. See for example Text::CSV_XS for how to write a (tab-separated) CSV file.

Modifying one key name in a hash does not make much sense to me and must be done copying the original to a new one and delete the original key, like in:

 $hash{'new_key'} = $hash{'key'};
 delete $hash{'key'};
[download]

L*

It may be a little unusual to construct hash keys that contain trailing whitespace, but putting that aside the challenge is that hash keys are immutable strings in Perl. Nevertheless, you can do what I think you're asking for like this:

$hash_sample_cds_genotype{"$_\t"}
    = delete $hash_sample_cds_genotype{$_} 
      for grep {m/^\Q$cds/} 
      keys %hash_sample_cds_genotype;
[download]

This is assuming that you wish to find all keys that start with a string that looks like $cds, and then append a tab to that key name. Since hash keys are immutable, you pretty much have to remove one element and add it back in under the new key name, which is what this code will do.

the answer is you make a new hash