substituting repeated regions in a long string

[jimineep]

Hi I have a long string in which I want to substitute repeated AT regions with N, ie ATATATATATATATATATAT NNNNNNNNNNNNNNNNNNNN I am having a problem doing this i have tried with tr/// which seems to mask all AT no matter how many repeats there are, and also with the s/// which behaves oddly. Enclosed is my latest attempt which prints the NNN sequence but prints it BEFORE the substituted sequence.

I realise I could use a for loop but I want to use this substitution for a sequence which is VERY long and I am sure there must be a better way

#!/usr/local/bin/perl

$testsequence=" ATGACGACTTATAGCGATGCTAGCATCTAGACTATATATATATATATATATATATATATATATATATATATATATATATATATATATATATATATATCGCTAGTAACGTAGTAGCTGTAGTAGCTGACTGATGCTGTAGTGACTGATGC";

$testsequence=~s/([A,T]{9,})/for($i=0;$i<length($1);$i++){ printf "N";}/e;

print $testsequence;
print "testsure";

 

[ishnid]

s/// doesn't behave oddly - you're just using it oddly ;-)

What about simply:

$testsequence =~ s/AT/NN/g;

That replaces all occurrences of "AT" with "NN"

Are you looking to replace all occurrences of the string "AT" or any combinations of the letters "A" and "T". I'm not entirely sure what you're attempting here.

 

[spookie]

If you want to replace both A & T with N no matter what sequence A & T are,

$testsequence =~ s/T/N/g ;
$testsequence =~ s/A/N/g ;

--------------------------------------------------------------------------
I never set a goal because u never know whats going to happen tommorow.

 

[jimineep]

I think u misunderstand me. I only want to replace AT with NN when AT is repeated 9 or more times(hence {9,})

ie in this string ATGCACAACAATATACGAATATATATATATATATATCGTATCGC

will become this

ATGCACAACAATATACGANNNNNNNNNNNNNNNNNNCGTATCGC

do u see the difference? I only want to replace repeat AT, not all instances of AT

 

[ishnid]

Much clearer. Try this:

$testsequence =~ s/((?:AT){9,})/'N' x length($1)/ge;

 

[jimineep]

gracias ishnid. What does the ?: in (?:AT) mean?

 

[ishnid]

It's a non-capturing group. Have a look at the "Non-capturing groupings" section of perlretut.