Question

我有两条长度相等的字符串，我需要比较一下。我想找到重叠基（。）和内部间隙（*）。以下是示例：

------ACTAAAAATACAAAAA--TTAGCCAGGCGTGGTGGCAC
-----TACTAAAAATACAAAAAAATTAGCCAGGTGTGGTGG---
      ................**.................

重叠次数= 33。内部间隙数= 2。

找到重叠次数没问题。但我有问题找到内部差距。以下是我目前的代码。它非常缓慢。原则上我需要计算数百万这样的对。

#!/usr/bin/perl -w
my $s1 = "------ACTAAAAATACAAAAA--TTAGCCAGGCGTGGTGGCAC";
my $s2 = "-----TACTAAAAATACAAAAAAATTAGCCAGGTGTGGTGG---";

print "$s1\n";
print "$s2\n";


my %base = ("A" => 1, "T" => 1, "C" => 1, "G" => 1);

my $ovlp_basecount = 0;
my $internal_gap = 0;

foreach my $si ( 0 .. length($s1)  ) {


    my $base1 = substr($s1,$si,1);
    my $base2 = substr($s2,$si,1);


    # Overlap
    if ( $base{$base1} && $base{$base2} ) {
        $ovlp_basecount++;
    }

    # Not sure how to compute internal gap

}


print "TOTAL OVERLAP BASE = $ovlp_basecount\n";
print "TOTAL Internal Gap \?\n";

请建议如何有效地找到内部差距和重叠。

Answer 1

您可以对字符串使用按位OR来查找一个字符串中与另一个字符串中的空白区域重叠的区域。此过程还具有通过将非重叠字符转换为小写来显示重叠的效果，从而使得重叠的查找非常简单：

#!/usr/bin/perl

use strict;
use warnings;

my $s1 = "------ACTAAAAATACAAAAA--TTAGCCAGGCGTGGTGGCAC";
my $s2 = "-----TACTAAAAATACAAAAAAATTAGCCAGGTGTGGTGG---";

$s1 =~ tr/-/\x20/;
$s2 =~ tr/-/\x20/;
my $or = $s1 | $s2;
(my $gap) = $or =~ m/^.*[ACTG]([actg]+)[ACTG].*$/;
(my $overlap = $or) =~ s/[^A-Z]//g;

print "s1:      '$s1'\n";
print "s2:      '$s2'\n";
print "OR:      '$or'\n";
printf "Gap:     '%s' (%d)\n", $gap,     length $gap;
printf "Overlap  '%s' (%d)\n", $overlap, length $overlap;

打印：

s1:      '      ACTAAAAATACAAAAA  TTAGCCAGGCGTGGTGGCAC'
s2:      '     TACTAAAAATACAAAAAAATTAGCCAGGTGTGGTGG   '
OR:      '     tACTAAAAATACAAAAAaaTTAGCCAGGWGTGGTGGcac'
Gap:     'aa' (2)
Overlap  'ACTAAAAATACAAAAATTAGCCAGGWGTGGTGG' (33)

有关字符串按位操作的更多信息：

http://teaching.idallen.com/cst8214/08w/notes/bit_operations.txt

Answer 2

假设间隙从不重叠，您可以使用正则表达式解决此问题。这是您s1的答案。

echo '------ACTAAAAATACAAAAA--TTAGCCAGGCGTGGTGGCAC' | perl -ne '$s = 0; foreach(/[GTAC](-+)[GTAC]/) { $s += length($1); } print "$s\n";'
2

在两个字符串中查找基本重叠计数和内部间隙

2 个答案: