Remove stickleback hack after web fix from Nick.

Header format column length fixed at 80 is too small for some GRCh38 seq_region names, forcing accession and version to span multiple lines.

if this is not the default for your mysql installation, characters will be bigger

Update eukaryota.pm

Added method documentation to explains why they get overridden

Overwrite transcript_names_from_gene method for all Ensembl Genomes species so it doesn't add '-201' suffix to the gene names.

we should not import our own annotation

these are being set when importing the LRGs

Self-syntenies: construct MLSS correctly.

Restore read coverage support

A warning more verbose, allow no alias, formatting

When retrieving the MLSS for self-synteny we must only specify the
species once so that we get one back.

Restore read_coverage data support removed in 75. Also support the has_coverage field and the downstream effects on rendered sequence of secondary strains - strains with no explicit read coverage data are represented in lowercase where there is no genotype call available.

code (e.g. Ensembl web) creates features without an adaptor.
In case of circular slices, return seq_region_start|end from the db only if
an adaptor is available.
If there is no adaptor it will try to do the normal calculation rather than die.

using internal_id as unique key instead

Other than some very minor formatting changes, the two changes here:
a) make a warning more verbose (I missed the meaning of the output)
b) allow the config file to have no alias value without causing an error.

Master