Extracting values from a complex and variable data structure to produce a report

Hello all --

I have an array of hashes that store data about apx. 4,000 genes.  A typical entry looks like this (some entries truncated with '...' for posting purposes):

BF2784 = {
  'name' => 'BF2784',
  'descr' => 'putative EPS related membrane protein',
  'start' => '3242515',
  'end' => '3244920',
  'ori' => 'pos',
  'bp' => '2409',
  'GC' => '56.3',
  'GeneID' => '3287061',
  'aa' => '802',
  'kDa' => '88.3',
  'GI' => '60682255',
  'groups' => {
    'CDD' => [
      'COG0455',
      'COG3206',
      'cd00550'
    ],
    'COG' => {
      'D' => 'Cell cycle control, mitosis and meiosis genes',
      'M' => 'Cell wall/membrane biogenesis genes'
    }
  },
  'links' => {
    'pep' => 'http://www.ncbi.nlm.nih.gov/entrez/...',
    'seq' => 'http://www.ncbi.nlm.nih.gov/entrez/...',
    'summary' => 'http://www.ncbi.nlm.nih.gov/entrez/...',
    'upstr' => 'http://www.ncbi.nlm.nih.gov/entrez/...'
  },
  'aka' => {
    'lec' => [
      'orf3_tsr19'
    ]
    'gb' => [
      'sigE'
    ],
  },
  'up_gap' => {
    'end' => '3242514',
    'size' => '13',
    'start' => '3242502'
  },
  'microarray' => {
    'SeqID' => 'BFRAG050600002693',
    'descr' => '2693|Bacteroides fragilis|0|506|CDS...'
    '102805' => {
      '0265_ML' => {
        'avg' => '959.5032',
        'block1' => '820.4571',
        'block2' => '1142.2980',
        'block3' => '915.7545'
      },
      '1394_EL' => {
        'avg' => '422.2764',
        'block1' => '448.5869',
        'block2' => '454.1586',
        'block3' => '364.0837'
      },
      '9343_EL' => {
        'avg' => '797.4852',
        'block1' => '753.0446',
        'block2' => '885.5215',
        'block3' => '753.8896'
      },
      '9343_ML' => {
        'avg' => '858.0540',
        'block1' => '933.8485',
        'block2' => '822.4420',
        'block3' => '817.8716'
      },
      'CrrD_ML' => {
        'avg' => '952.3332',
        'block1' => '1000.5565',
        'block2' => '949.4948',
        'block3' => '906.9484'
      }
    },
    '121905' => {
      '9343_ML' => {
        'avg' => '976.8530',
        'block1' => '1053.0826',
        'block2' => '930.7049',
        'block3' => '946.7716'
      },
      'ddUngD' => {
        'avg' => '852.5260',
        'block1' => '851.9713',
        'block2' => '823.1842',
        'block3' => '882.4226'
      },
      'mpi_mut44' => {
        'avg' => '1295.1745',
        'block1' => '1367.4020',
        'block2' => '1229.0144',
        'block3' => '1289.1070'
      },
      'mpi_mut8' => {
        'avg' => '1126.2450',
        'block1' => '1115.6544',
        'block2' => '1093.3422',
        'block3' => '1169.7385'
      },
      'tsr19_M1' => {
        'avg' => '1895.5840',
        'block1' => '1916.8111',
        'block2' => '1798.7082',
        'block3' => '1971.2327'
      },
      'tsr19_M3' => {
        'avg' => '1249.8808',
        'block1' => '1215.6576',
        'block2' => '1281.3577',
        'block3' => '1252.6272'
      }
    },
  },
};

These data are stored on disk in Storable format, and retrieved by:

  use Storable qw(store retrieve);
  my $data = retrieve("master_9343.db");

The entry above is from a hash in the $data->[0] array, e.g.:

use Data::Dumper;

open (DD, ">BF2784_dump.txt") or die;
print DD Dumper($data->[0]{BF2784});

This arrangement works great if I'm looking up a particular attribute of the gene, or a small set of attributes:

print "$gene begins at $data->[0]{$gene}{start} and ends at $data->[0]{$gene}{end}\n";

but what I want to do now is provide a program to display *all* stored data about the gene.  Some genes do not have all the entries shown above (for example, there may be only one name for a particular gene, thus neither $data->[0]{$gene}{aka}{lec} nor $data->[0]{$gene}{aka}{gb} will exist for that value of $gene, but either or both might exist for another value of $gene).  This is true for several of the structures ($data->[0]{$gene}{groups}, $data->[0]{$gene}{up_gap}, etc.).

So, my goal is to write a command line program that, when provided the name of a gene, prints out a report containing all the data accumulated for that gene.  For now, I'm going to output it to a text file using Perl's Report formats, but I might eventually try for a Perl/Tk version.

What is the best way to iterate over such a variable structure to dump all the data, ignoring values that don't exist?  I could of course set variables for each possible entry, testing first if it exists:

if (exists $data->[0]{$gene}{aka}{lec}) {
   my $aka = join (", ", @{$data->[0]{$gene}{aka}{lec}});
}

if (exists $data->[0]{$gene}{groups}{CDD}) {
   my $cdd = join (", ", @{$data->[0]{$gene}{groups}{CDD}});
}

etc., but this brute force approach seems tedious and wasteful.  Does anyone have any suggestions for an efficient way to do this, such that I wind up with a collection of variables suitable to pass to a report format subfunction?

Thanks --

MikeStart Free Trial