finding sequence

yueli711 has asked for the wisdom of the Perl Monks concerning the following question:

Hello, I want to find the sequences in tmp01 which are containing in tmp02, then print all of them out. Tab seperated tmp02 file. Thanks in advance!

 tmp01

ATCCCACCGCTGCCACCA
ACCCTGCTCGCTGCGCCA
TCCCCGGCACCTCCACCA
TCCCCGGCATCTCCACCA
ATCCTGCCGACTACGCCA
TCGATTCCCGGCCCATGCACCA
TCGATTCCCGGCCAACGCACCA
GTCCCACCAGAGTCGCCA
ACCCCACTCCTGGTACCA
GTCCCTTCGTGGTCGCCA

tmp02

AACCCCATCCCACCGCTGCCACCA    1
AACCCCATCCTCGTCGCC    1
AACCCCATGAAATAAGAG    2
AACCCCATGATCAGGACAAG    1
AACCCCATTAAAAAATGG    1
AACTGGATTCTCTGAAATCCCACCGCTGCCACCA    1
AACTGGATTGTCTGTTTGT    1
AACTGGCAAGTTCAGGCATG    1
AACTGGCACACACAACC    1
AACTGGCACACACAACCT    1
[download]


open(IN1,"tmp01") || die "Cannot open this file";

@lines1 = <IN1>;

open(IN2,"tmp02") || die "Cannot open this file";

@lines2 = <IN2>;


open(OUT,">tmp03") || die "Cannot open this file";

for $item1(@lines1){

    chomp $item1;


    #print OUT $item1,"\t";

@tmp1=split(/\t+/, $item1);


for $item2(@lines2){

    chomp $item2;

@tmp2=split(/\t+/, $item2);

        if ($tmp1[0] =~m/ *$tmp2[0]*){

            print OUT $item1,"\t",$item2;

            #last;    

        }

    $i++
}


print OUT "\n";


}


close(IN1);

close(IN2);

close(OUT);
[download]

Comment on finding sequence Select or Download Code

Replies are listed 'Best First'.
Re: finding sequence by haukex (Archbishop) on Jun 22, 2018 at 19:40 UTC
If I fix your regular expression (see perlretut), then note that `$tmp1[0] =~ /\Q$tmp2[0]\E/` means "is the string `$tmp2[0]` contained in `$tmp1[0]`?". However, just as an example, `$tmp1[0]` may be `"ATCCCACCGCTGCCACCA"`, while `$tmp2[0]` may be `"AACCCCATCCCACCGCTGCCACCA"` - so as you might be able to tell, you've got your condition reversed. Some better variable naming would really help here! You haven't shown your expected output, so I can't give any advice there other than to have a look at How do I post a question effectively?, SSCCE, and I know what I mean. Why don't you? A few general tips: Use strict and warnings (you've been told this twice before), and use the newer, three-argument form of open and lexical filehandles, as in `open(my $infh1, '<', $filename) or die "Cannot open $filename: $!";` Update: By the way, you could build a single regex out of all of the search strings - I wrote a tutorial on this at Building Regex Alternations Dynamically.	[reply] [d/l] [select]
Re^2: finding sequence by yueli711 (Sexton) on Jun 22, 2018 at 20:19 UTC
It really solve my problem! Thanks!	[reply]
Re: finding sequence by hippo (Archbishop) on Jun 22, 2018 at 18:58 UTC
`if ($tmp1[0] =~m/ $tmp2[0]){` Your match pattern isn't terminated. Perl tells you this: `$ perl 1217240.pl Search pattern not terminated at 1217240.pl line 29.` [download] Why the random indentation? It makes your code so hard to read. I'm also mildly intrigued as to why every other line is blank.	[reply] [d/l] [select]
Re: finding sequence by Laurent_R (Canon) on Jun 23, 2018 at 09:03 UTC
Since you're looking for exact matches, I believe that using the index function might be more efficient than regex matching. In general, `index` is faster than regexes for exact matches, but, in that specific case, building a large pattern with all the search strings, as shown by tybalt89, might actually end up to be faster. You might also consider to store `fh2` into a large string (joined by an appropriate separator), this is also likely to be faster than looping over the content of `fh2`.	[reply] [d/l] [select]
Re: finding sequence by tybalt89 (Monsignor) on Jun 22, 2018 at 23:09 UTC
#!/usr/bin/perl # https://perlmonks.org/?node_id=1217240 use strict; use warnings; open my $patterns, '<', \<<END; ATCCCACCGCTGCCACCA ACCCTGCTCGCTGCGCCA TCCCCGGCACCTCCACCA TCCCCGGCATCTCCACCA ATCCTGCCGACTACGCCA TCGATTCCCGGCCCATGCACCA TCGATTCCCGGCCAACGCACCA GTCCCACCAGAGTCGCCA ACCCCACTCCTGGTACCA GTCCCTTCGTGGTCGCCA END open my $fh, '<', \<<END; AACCCCATCCCACCGCTGCCACCA 1 AACCCCATCCTCGTCGCC 1 AACCCCATGAAATAAGAG 2 AACCCCATGATCAGGACAAG 1 AACCCCATTAAAAAATGG 1 AACTGGATTCTCTGAAATCCCACCGCTGCCACCA 1 AACTGGATTGTCTGTTTGT 1 AACTGGCAAGTTCAGGCATG 1 AACTGGCACACACAACC 1 AACTGGCACACACAACCT 1 END my $seqs = join '\|', map tr/ACGT//cdr, <$patterns>; my $match = qr/^\w*($seqs)/; /$match/ and print "$1\t$_" while <$fh>; [download] Outputs: `ATCCCACCGCTGCCACCA AACCCCATCCCACCGCTGCCACCA 1 ATCCCACCGCTGCCACCA AACTGGATTCTCTGAAATCCCACCGCTGCCACCA 1` [download]	[reply] [d/l] [select]