sqm2csv.pl

#!/usr/bin/env perl
# Sam Shepard - 2013
# Square distance matrix to CSV format

use Getopt::Long;
GetOptions(
            'csv|C=s'       => \$csv,
            'use-index|I'   => \$useIndex,
            'dist-name|N=s' => \$distName,
            'set-name|S=s'  => \$sampleSet,
            'vsX|V=s'       => \$vsX,
            'full-matrix|F' => \$fullMatrix,
            'vs-list|L=s'   => \$vsList,
            'remove-refs'   => \$removeRefs,
            'delim|D=s'     => \$delim
);

# VERSUS X
# Annotation field
# Strip annots from ID
if ( scalar(@ARGV) != 1 ) {
    $message = "Usage:\n\tperl $0 <square_matrix.tab>\n";
    $message .= "\t\t-D|--delim\t\t\tDelimiter, default is TAB.\n";
    $message .= "\t\t-C|--csv <file.csv>\t\tA different CSV to use for ID-pair order.\n";
    $message .= "\t\t-I|--use-index\t\t\tUse numerical indices rather than the IDs.\n";
    $message .= "\t\t-N|--dist-name <NAME>\t\tRename the distance header.\n";
    $message .= "\t\t-S|--set-name\t\t\tAdd a column for the set name.\n";
    $message .= "\t\t-V|--vsX <ID>\t\t\tOnly print versus ID.\n";
    $message .= "\t\t-L|--vs-list <ID1[,ID2,..]>\tOne column is used per list reference (sqm subset).\n";
    $message .= "\t\t-R|--remove-refs\t\tGiven above, remove reference rows.\n";
    die( $message . "\n" );
}

if ( !defined($sampleSet) ) {
    $sampleSet = '';
} else {
    $sampleSet = "," . $sampleSet;
}

if ( defined($delim) ) {
    $d = $delim;
} else {
    $d = "\t";
}

open( IN, '<', $ARGV[0] ) or die("Cannot open $ARGV[0].\n");
$/ = "\n";
$i = 0;
while ( $line = <IN> ) {
    chomp($line);
    @values         = split( "\t", $line );
    $id             = shift(@values);
    $distHash{$id}  = [@values];
    $indexByID{$id} = $i;
    @IDs[$i]        = $id;
    $i++;
}
close(IN);
$N = $i;
if ( !defined($distName) ) { $distName = 'dist'; }

# process vs List
if ( defined($vsList) ) {
    %vsID = ();
    foreach $id ( split( ',', $vsList ) ) {
        if ( defined( $indexByID{$id} ) ) {
            $vsID{$id} = $indexByID{$id};
        } else {
            print STDERR "Invalid reference, $id. Skipping.\n";
        }
    }
    $vsList = 1;
} else {
    $vsList = 0;
}

if ( !$vsList ) {
    if ( defined($vsX) ) {
        print "vs $vsX$d$distName\n";
    } else {
        print "id1${d}id2${d}$distName\n";
    }
}
if ( defined($csv) ) {
    open( IN, '<', $csv ) or die("Cannot open $csv.\n");
    $header = <IN>;
    while ( $line = <IN> ) {
        chomp($line);
        @values = split( ',', $line );
        $id1    = $values[0];
        $id2    = $values[1];

        if ( !defined($vsX) ) {
            print $id1, $d, $id2, $d, $distHash{$id1}[$indexByID{$id2}], $sampleSet, "\n";
        } elsif ( $id1 eq $vsX ) {
            print $id2, $d, $distHash{$id1}[$indexByID{$id2}], $sampleSet, "\n";
        } elsif ( $id2 eq $vsX ) {
            print $id1, $d, $distHash{$id1}[$indexByID{$id2}], $sampleSet, "\n";
        }
    }
    close(IN);
} else {
    if ($vsList) {
        @columns = sort { $a <=> $b } values(%vsID);
        print 'ID';
        if   ($distName) { $suffix = '_' . $distName; }
        else             { $suffix = ''; }
        foreach $col (@columns) { print $d, $IDs[$col], $suffix; }
        print "\n";
        for ( $i = 0; $i < $N; $i++ ) {
            if ( $removeRefs && defined( $vsID{ $IDs[$i] } ) ) { next; }

            print $IDs[$i];
            foreach $col (@columns) {
                print $d, $distHash{ $IDs[$i] }[$col];
            }
            print "\n";
        }
    } elsif ($fullMatrix) {
        for ( $i = 0; $i < $N; $i++ ) {
            for ( $j = 0; $j < $N; $j++ ) {
                if ( !defined($vsX) ) {
                    if ($useIndex) {
                        print $i, $d, $j, $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    } else {
                        print $IDs[$i], $d, $IDs[$j], $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    }
                } elsif ( $IDs[$i] eq $vsX ) {
                    if ($useIndex) {
                        print $j, $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    } else {
                        print $IDs[$j], $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    }
                } elsif ( $IDs[$j] eq $vsX ) {
                    if ($useIndex) {
                        print $i, $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    } else {
                        print $IDs[$i], $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    }
                }
            }
        }
    } else {
        for ( $i = 1; $i < $N; $i++ ) {
            for ( $j = 0; $j < $i; $j++ ) {
                if ( !defined($vsX) ) {
                    if ($useIndex) {
                        print $i, $d, $j, $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    } else {
                        print $IDs[$i], $d, $IDs[$j], $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    }
                } elsif ( $IDs[$i] eq $vsX ) {
                    if ($useIndex) {
                        print $j, $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    } else {
                        print $IDs[$j], $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    }
                } elsif ( $IDs[$j] eq $vsX ) {
                    if ($useIndex) {
                        print $i, $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    } else {
                        print $IDs[$i], $d, $distHash{ $IDs[$i] }[$j], $sampleSet, "\n";
                    }
                }
            }
        }
    }
}