parse_genbank.pl

#! /usr/local/bin/perl -w

# Homemade Genbank report parser using regular expressions.
# Once desired data is captured, it can be printed in any format.
# WI Biocomputing course - Unix and Programming Skills for Biologists - March 2003

$gb_report = "genbank_sample.txt";

open (GB, $gb_report) || die "cannot open $gb_report for reading: $!";

# Flag for multiline translation; 1 means translation "in preogress"  
$trans = 0;

while (<GB>)
{
   if (/(LOCUS\s*)(\w*)(.*)/)
   {
      print "Locus: $2\n";
   }
   elsif (/(VERSION.*GI:)(\d*)/)
   {
      print "GI: $2\n";
   }
   elsif (/(DEFINITION\s*)(.*)(\.)/)
   {
      print "Sequence name: $2\n";
   }
   elsif (/(ORGANISM\s*)(.*)/)
   {
      print "Organism: $2\n";
   }
   elsif(/(gene)(\s*)(\d*)(\.\.)(\d*)/)
   {
      print "Gene length: $5\n";
   }
   elsif (/(CDS\s*)(\d*)(\.\.)(\d*)/)
   # ex: CDS             357..1541
   {
      $cds_start = $2;
      $cds_end = $4;
      print "CDS: $cds_start - $cds_end\n";
   }
   elsif (/(\/translation=")(.*)/)   # protein product begins
   {
      print "Translation: ";
      $protein = $2;
      $trans = 1;
   }
   elsif ($trans)   # translation still going on
   {
      if (!/"/)   # no terminal quote; translation continues
      {
         $protein .= $_;
      }
      elsif (/(.*)(")/)   # terminal quote; end of translation
      {
         $protein .= $1;
         $protein =~ s/\s*//g;
         print "$protein\n";
         $trans = 0;
      }
      else
      {
         print "Problems: end of translation product not found.\n";
      }
   }
   else
   {
      # Skip this data
   }
}