GetCUISimilarity.pl

#!/usr/bin/perl
#
# Author:         Linus Kohl
# E-Mail:         kohl@munichresearch.com
# Org:            MunichResearch
#
use strict;
use UMLS::Interface;
use UMLS::Similarity::lch;
use UMLS::Similarity::path;
use Parse::CSV;
use Text::CSV;
use Try::Tiny;
use Term::ProgressBar;

my $filenameInput  = $ARGV[0];
my $filenameOutput = $ARGV[1];

# Get number of lines
my $numPairings = `wc -l < $filenameInput`;

# Open files
open( my $ioIn, '<:encoding(UTF-8)', $filenameInput )
  or die "Failed to open file '$filenameInput': $!";

open( my $ioOut, ">:encoding(utf8)", $filenameOutput )
  or die "Failed to create output file: $!";

# CSV input
my $csvIn = Parse::CSV->new( handle => $ioIn, sep_char => ',' );

# CSV output
my $csvOut = Text::CSV->new();
$csvOut->eol("\r\n");
$csvOut->column_names( 'cuid1', 'cuid2', 'lch', 'path' );

# Initialize UMLS interfaces
my $umls = UMLS::Interface->new();
die "Unable to create UMLS::Interface object.\n" if ( !$umls );

my $path = UMLS::Similarity::path->new($umls);
die "Unable to create Path Similarity object.\n" if ( !$path );

my $lch = UMLS::Similarity::lch->new($umls);
die "Unable to create LCH Similarity object.\n" if ( !$lch );

my $progress = Term::ProgressBar->new($numPairings);

while ( my $rowRef = $csvIn->fetch ) {

    # uncomment if file contains header row
    # next if $objects->row==1;
    my @row = @{$rowRef};
    try {
        # find coresponding term by CUI
        my ( $id, $cui1, $cui2 ) = @row;
        # get distances
        my $pathValue = $path->getRelatedness( $cui1, $cui2 );
        my $lchValue = $lch->getRelatedness( $cui1, $cui2 );

        # write values to file
        $csvOut->print( $ioOut, [ $cui1, $cui2, $lchValue, $pathValue ] );
    }
    catch {
        # warn "Warning: '$_'";
    }
    finally {
        $progress->update($_);
    };
}

close $ioOut or die $!;
close $ioIn  or die $!;