circular slices. In this case, do not use seq_region_start|end to compute min_start|end as the
feature might be crossing the origin of replication and the comparisons to determine min_start
and max_end might then be nonsense.

still need to remove manually any mappings which would have been added before for that patch

circular slices. In this case, do not use seq_region_start|end to compute min_start|end as the
feature might be crossing the origin of replication and the comparisons to determine min_start
and max_end might then be nonsense.

use production database to fetch biotype groups

[ENSCORESW-840]. seq region boundaries are fetched from the db (for slices stored in the db) in case
of circular chromosomes. This is because it is impossible to recompute the seq_region_start|end for
a feature in a circular slice since the transformation operated in method _objs_from_sth in the
corresponding adaptor is not invertible.
For some of the features (e.g. those in the Bio::EnsEMBL::FeaturePair hierarchy), we can't query
the db and the methods seq_region_start|end return undef even if the feature is on a slice.

all modules take the same arguments and have the same behaviour
the only differences should be related to the columns specific to a given feature

add splicingEvent tests for basic methods

* hotfix/comparadocumentation:
  [ENSCORESW-841] - Doxygen output. Make PerlFilter ignore additional packages in a module, to prevent confusion in Compara codebase.

[ENSCORESW-841] - Doxygen output. Make PerlFilter ignore additional packages in a module, to prevent confusion in Compara codebase.

We want people to give us bugfixes or cool features which help
everyone. So we're making the rules to contributing easier.

restore changes from c981f63d

lost during pull request #4 to f9a90829

added extra_data/external_data column to insert for storing DnaAlignFeatures

use code directly in get_attrib method

A fix for the fix. Originally we started with passing through Slice
for the aligned feature but this is wrong. We need to assert the
hit feature Slice which we do not have. This change reverts the
_check_start_end_strand() method back to its original functionality.

Revert "transcript counts can now be fetched using transcript_cnt and transcript_acnt"

This reverts commit 3d9c5ad7.

* dev:
  transcript table in patch database now contains data genome_container test module updated to reflect the change

genome_container test module updated to reflect the change

* dev:
  there can be several statistics for prediction_count

Introducing a new version of our GTF dumper. Now supports
gene and transcript features, seleno markup, ccds xrefs and
a new readme.

* dev:
  corrected argument for structural_variation count