[Bioperl-l] Get nucleic CDS sequence from Genbank files using spliced_seq

[Johann PELLET]

Hi,
I have two questions:
First, I have problems for example with this genbank entry: NC_008210. 
Indeed, for one CDS, the location is : 
join(161990..162784,complement(88222..88806),complement(86666..87448))

When I parse a genbank file with this entry.

my $seq_in  = Bio::SeqIO->new( -format => 'genbank',
                               -file => $input_file);

while( my $seq = $seq_in->next_seq() ) {
    my @features = $seq->get_SeqFeatures();
    for ( my $i =0; $i < scalar @features; $i++ ){
        my $feat = @features[$i];
         if ( $feat->primary_tag eq 'CDS' ){


             my $seq_CDS_obj=$cds->spliced_seq( -nosort => 0);
             my $seq_CDS=$seq_CDS_obj->seq;


I have this error:
Can't call method "isa" without a package or object reference at ....

Secondly, When location is like: 
join(complement(AY421753.1:1..6),complement(3813..5699)) I know that we 
must use spliced_seq with the argument db, but it's not working.
      my $seq_CDS_obj=$cds->spliced_seq( -db => "genbank");
How can we valid a Bio::DB::RandomAccessI?

Thanks

-- 
Johann Pellet
phone (work): +33(0)4 37282352
E-mail:    pellet at cervi-lyon.inserm.fr
Centre d'Etudes et de Recherche en Virologie et Immunologie
INSERM U503
21, Avenue Tony Garnier
69365 Lyon cedex 07 France