GitLab

to difference in seq_region names.

Previous change backed out, as not in all cases 1bp slice start = seq_region_start or slice end = seq_region_end.

Adding a more generic URI parser to replace the numerous places where people parse URIs in API code (mostly to get DB params)

bringing in /codon_start to represent offsets in the CDS based upon phase. The translation from phase to codon_start is non-obvious & has been documented.

64 to 255 characters.  Requested by Nick Langridge.

Also, I got a logic test back to front (fixed).

result in infinite recursion.

Due to a bug in MySQL we need to drop the unique index, change the field to varchar then reaply the index otherwise latin1 encodings can cause a constraint error to occur. This only affects the conversion of enum to varchar

and mark the canonical one as canonical.  Solves issue raised
by Will McLaren where none of the transcripts returned from
$gene->get_all_Transcripts() would be marked as canonical.

($use_projection_code at top of module) for debugging.

New methods 'get_all_StructuralVariationFeatures' and 'get_all_CopyNumberVariantProbeFeatures' to replace 'get_all_StructuralVariations' and 'get_all_CopyNumberVariantProbes' (deprecated)

Force the loading of external synonyms before getting the code to store them. Otherwise we incorrectly retrieve external synonyms from the source database with a newly assigned ID

fixes when identifying other stops -  1) ribosomal frameshift ignored, 2) stop codon being automatically annotated as a seleno and 3) non numerical values in comments

source are on different seq_regions.  Also check the exonerate score
threshold before adding the score to the exonerate scoring matrix.

Values will be tested in healthchecks.