get_synonyms.py fails to infer the right URL

[priesgo]

Dear fusion catchers,

While building the indexes for fusion catcher we stumbled upon an issue that may be derived from a change in the FTP structure that break some assumptions.

When the organism is homo_sapiens there are two subfolders homo_sapiens_core_112_37 and homo_sapiens_core_112_38 under ftp.ensembl.org/pub/current_mysql/

In this line https://github.com/ndaniel/fusioncatcher/blob/master/bin/get_synonyms.py#L139 if organism_dir and len(organism_dir) == 1: it is assumed that there is only one folder and hence it fails to download the required resources.

Possible solution:

• Just grab the last one?
• Specify the genome version explicitly?

Thanks!

[lynnhlee]

Bumping this to the top - encountered the same error trying to build reference files for 1.33...