use warnings;
use Getopt::Std;
-my $version = '0.3.2 (r321)';
+my $version = '0.3.3';
&usage if (@ARGV < 1);
my $command = shift(@ARGV);
-my %func = (showALEN=>\&showALEN, pileup2fq=>\&pileup2fq, varFilter=>\&varFilter, unique=>\&unique);
+my %func = (showALEN=>\&showALEN, pileup2fq=>\&pileup2fq, varFilter=>\&varFilter,
+ unique=>\&unique, uniqcmp=>\&uniqcmp, sra2hdr=>\&sra2hdr, sam2fq=>\&sam2fq);
die("Unknown command \"$command\".\n") if (!defined($func{$command}));
&{$func{$command}};
die(qq/Usage: samtools.pl showALEN <in.sam>\n/) if (@ARGV == 0 && -t STDIN);
while (<>) {
my @t = split;
+ next if (/^\@/ || @t < 11);
my $l = 0;
$_ = $t[5];
- s/(\d+)[SMI]/$l+=$1/eg;
+ s/(\d+)[MI]/$l+=$1/eg;
print join("\t", @t[0..5]), "\t$l\t", join("\t", @t[6..$#t]), "\n";
}
}
# varFilter
#
+#
+# Filtration code:
+#
+# d low depth
+# D high depth
+# W too many SNPs in a window (SNP only)
+# G close to a high-quality indel (SNP only)
+# Q low RMS mapping quality (SNP only)
+# g close to another indel with higher quality (indel only)
+
sub varFilter {
my %opts = (d=>3, D=>100, l=>30, Q=>25, q=>10, G=>25, s=>100, w=>10, W=>10, N=>2, p=>undef);
- getopts('pd:D:l:Q:w:W:N:G:', \%opts);
+ getopts('pq:d:D:l:Q:w:W:N:G:', \%opts);
die(qq/
Usage: samtools.pl varFilter [options] <in.cns-pileup>
my @staging; # (indel_filtering_score, flt_tag)
while (<>) {
my @t = split;
- next if ($t[2] eq $t[3] || $t[3] eq '*/*'); # skip non-var sites
+ next if (uc($t[2]) eq uc($t[3]) || $t[3] eq '*/*'); # skip non-var sites
# clear the out-of-range elements
while (@staging) {
last if ($staging[0][2] eq $t[0] && $staging[0][3] + $max_dist >= $t[1]);
}
}
+#
+# sam2fq
+#
+
+sub sam2fq {
+ my %opts = (n=>20, p=>'');
+ getopts('n:p:', \%opts);
+ die("Usage: samtools.pl sam2fq [-n 20] [-p <prefix>] <inp.sam>\n") if (@ARGV == 0 && -t STDIN);
+ if ($opts{p} && $opts{n} > 1) {
+ my $pre = $opts{p};
+ my @fh;
+ for (0 .. $opts{n}-1) {
+ open($fh[$_], sprintf("| gzip > $pre.%.3d.fq.gz", $_)) || die;
+ }
+ my $i = 0;
+ while (<>) {
+ next if (/^@/);
+ chomp;
+ my @t = split("\t");
+ next if ($t[9] eq '*');
+ my ($name, $seq, $qual);
+ if ($t[1] & 16) { # reverse strand
+ $seq = reverse($t[9]);
+ $qual = reverse($t[10]);
+ $seq =~ tr/ACGTacgt/TGCAtgca/;
+ } else {
+ ($seq, $qual) = @t[9,10];
+ }
+ $name = $t[0];
+ $name .= "/1" if ($t[1] & 0x40);
+ $name .= "/2" if ($t[1] & 0x80);
+ print {$fh[$i]} "\@$name\n$seq\n";
+ if ($qual ne '*') {
+ print {$fh[$i]} "+\n$qual\n";
+ }
+ $i = 0 if (++$i == $opts{n});
+ }
+ close($fh[$_]) for (0 .. $opts{n}-1);
+ } else {
+ die("To be implemented.\n");
+ }
+}
+
+#
+# sra2hdr
+#
+
+# This subroutine does not use an XML parser. It requires that the SRA
+# XML files are properly formated.
+sub sra2hdr {
+ my %opts = ();
+ getopts('', \%opts);
+ die("Usage: samtools.pl sra2hdr <SRA.prefix>\n") if (@ARGV == 0);
+ my $pre = $ARGV[0];
+ my $fh;
+ # read sample
+ my $sample = 'UNKNOWN';
+ open($fh, "$pre.sample.xml") || die;
+ while (<$fh>) {
+ $sample = $1 if (/<SAMPLE.*alias="([^"]+)"/i);
+ }
+ close($fh);
+ # read experiment
+ my (%exp2lib, $exp);
+ open($fh, "$pre.experiment.xml") || die;
+ while (<$fh>) {
+ if (/<EXPERIMENT.*accession="([^\s"]+)"/i) {
+ $exp = $1;
+ } elsif (/<LIBRARY_NAME>\s*(\S+)\s*<\/LIBRARY_NAME>/i) {
+ $exp2lib{$exp} = $1;
+ }
+ }
+ close($fh);
+ # read run
+ my ($run, @fn);
+ open($fh, "$pre.run.xml") || die;
+ while (<$fh>) {
+ if (/<RUN.*accession="([^\s"]+)"/i) {
+ $run = $1; @fn = ();
+ } elsif (/<EXPERIMENT_REF.*accession="([^\s"]+)"/i) {
+ print "\@RG\tID:$run\tSM:$sample\tLB:$exp2lib{$1}\n";
+ } elsif (/<FILE.*filename="([^\s"]+)"/i) {
+ push(@fn, $1);
+ } elsif (/<\/RUN>/i) {
+ if (@fn == 1) {
+ print STDERR "$fn[0]\t$run\n";
+ } else {
+ for (0 .. $#fn) {
+ print STDERR "$fn[$_]\t$run", "_", $_+1, "\n";
+ }
+ }
+ }
+ }
+ close($fh);
+}
+
#
# unique
#
sub unique {
- my %opts = (f=>5.0, q=>5, r=>2, a=>1, b=>3);
- getopts('f:', \%opts);
+ my %opts = (f=>250.0, q=>5, r=>2, a=>1, b=>3);
+ getopts('Qf:q:r:a:b:m', \%opts);
die("Usage: samtools.pl unique [-f $opts{f}] <in.sam>\n") if (@ARGV == 0 && -t STDIN);
my $last = '';
+ my $recal_Q = !defined($opts{Q});
+ my $multi_only = defined($opts{m});
my @a;
while (<>) {
my $score = -1;
print $_ if (/^\@/);
$score = $1 if (/AS:i:(\d+)/);
my @t = split("\t");
+ next if (@t < 11);
if ($score < 0) { # AS tag is unavailable
my $cigar = $t[5];
my ($mm, $go, $ge) = (0, 0, 0);
$cigar =~ s/(\d+)M/$mm+=$1/eg;
$score = $mm * $opts{a} - $go * $opts{q} - $ge * $opts{r}; # no mismatches...
}
- $score = 0 if ($score < 0);
+ $score = 1 if ($score < 1);
if ($t[0] ne $last) {
- &unique_aux(\@a, $opts{f}) if (@a);
+ &unique_aux(\@a, $opts{f}, $recal_Q, $multi_only) if (@a);
$last = $t[0];
}
push(@a, [$score, \@t]);
}
- &unique_aux(\@a, $opts{f}) if (@a);
+ &unique_aux(\@a, $opts{f}, $recal_Q, $multi_only) if (@a);
}
sub unique_aux {
- my ($a, $fac) = @_;
- my ($max, $max2, $max_i) = (-1, -1, -1);
+ my ($a, $fac, $is_recal, $multi_only) = @_;
+ my ($max, $max2, $max_i) = (0, 0, -1);
for (my $i = 0; $i < @$a; ++$i) {
if ($a->[$i][0] > $max) {
$max2 = $max; $max = $a->[$i][0]; $max_i = $i;
$max2 = $a->[$i][0];
}
}
- my $q = int($fac * ($max - $max2) + .499);
- $a->[$max_i][1][4] = $q < 250? $q : 250;
+ if ($is_recal) {
+ if (!$multi_only || @$a > 1) {
+ my $q = int($fac * ($max - $max2) / $max + .499);
+ $q = 250 if ($q > 250);
+ $a->[$max_i][1][4] = $q < 250? $q : 250;
+ }
+ }
print join("\t", @{$a->[$max_i][1]});
@$a = ();
}
+#
+# uniqcmp: compare two SAM files
+#
+
+sub uniqcmp {
+ my %opts = (q=>10, s=>100);
+ getopts('pq:s:', \%opts);
+ die("Usage: samtools.pl uniqcmp <in1.sam> <in2.sam>\n") if (@ARGV < 2);
+ my ($fh, %a);
+ warn("[uniqcmp] read the first file...\n");
+ &uniqcmp_aux($ARGV[0], \%a, 0);
+ warn("[uniqcmp] read the second file...\n");
+ &uniqcmp_aux($ARGV[1], \%a, 1);
+ warn("[uniqcmp] stats...\n");
+ my @cnt;
+ $cnt[$_] = 0 for (0..9);
+ for my $x (keys %a) {
+ my $p = $a{$x};
+ my $z;
+ if (defined($p->[0]) && defined($p->[1])) {
+ $z = ($p->[0][0] == $p->[1][0] && $p->[0][1] eq $p->[1][1] && abs($p->[0][2] - $p->[1][2]) < $opts{s})? 0 : 1;
+ if ($p->[0][3] >= $opts{q} && $p->[1][3] >= $opts{q}) {
+ ++$cnt[$z*3+0];
+ } elsif ($p->[0][3] >= $opts{q}) {
+ ++$cnt[$z*3+1];
+ } elsif ($p->[1][3] >= $opts{q}) {
+ ++$cnt[$z*3+2];
+ }
+ print STDERR "$x\t$p->[0][1]:$p->[0][2]\t$p->[0][3]\t$p->[0][4]\t$p->[1][1]:$p->[1][2]\t$p->[1][3]\t$p->[1][4]\t",
+ $p->[0][5]-$p->[1][5], "\n" if ($z && defined($opts{p}) && ($p->[0][3] >= $opts{q} || $p->[1][3] >= $opts{q}));
+ } elsif (defined($p->[0])) {
+ ++$cnt[$p->[0][3]>=$opts{q}? 6 : 7];
+ print STDERR "$x\t$p->[0][1]:$p->[0][2]\t$p->[0][3]\t$p->[0][4]\t*\t0\t*\t",
+ $p->[0][5], "\n" if (defined($opts{p}) && $p->[0][3] >= $opts{q});
+ } else {
+ print STDERR "$x\t*\t0\t*\t$p->[1][1]:$p->[1][2]\t$p->[1][3]\t$p->[1][4]\t",
+ -$p->[1][5], "\n" if (defined($opts{p}) && $p->[1][3] >= $opts{q});
+ ++$cnt[$p->[1][3]>=$opts{q}? 8 : 9];
+ }
+ }
+ print "Consistent (high, high): $cnt[0]\n";
+ print "Consistent (high, low ): $cnt[1]\n";
+ print "Consistent (low , high): $cnt[2]\n";
+ print "Inconsistent (high, high): $cnt[3]\n";
+ print "Inconsistent (high, low ): $cnt[4]\n";
+ print "Inconsistent (low , high): $cnt[5]\n";
+ print "Second missing (high): $cnt[6]\n";
+ print "Second missing (low ): $cnt[7]\n";
+ print "First missing (high): $cnt[8]\n";
+ print "First missing (low ): $cnt[9]\n";
+}
+
+sub uniqcmp_aux {
+ my ($fn, $a, $which) = @_;
+ my $fh;
+ $fn = "samtools view $fn |" if ($fn =~ /\.bam/);
+ open($fh, $fn) || die;
+ while (<$fh>) {
+ my @t = split;
+ next if (@t < 11);
+# my $l = ($t[5] =~ /^(\d+)S/)? $1 : 0;
+ my $l = 0;
+ my ($x, $nm) = (0, 0);
+ $nm = $1 if (/NM:i:(\d+)/);
+ $_ = $t[5];
+ s/(\d+)[MI]/$x+=$1/eg;
+ @{$a->{$t[0]}[$which]} = (($t[1]&0x10)? 1 : 0, $t[2], $t[3]-$l, $t[4], "$x:$nm", $x - 4 * $nm);
+ }
+ close($fh);
+}
+
#
# Usage
#
projects / samtools.git / blobdiff