Fixed the ensembl transcript gene_id and msigdb geneSet publication

[evergreen700]

Unsure if the publication ID is the format you wanted but at least it's attached to the record now. I'll probably look at another record with publications next and make sure the numbers are able to match each other.

[evergreen700]

Pubmed:

• Changed the transform.yaml to be more flexible with different absrtract formats
• Changed the download files to work a little bit better. There is still a snakemake bug in there but that would probably be better addressed within the lathe source code.

[evergreen700]

Cellosarus:

• Obo reader modified slightly to capture species name but shouldn't affect execution of sifter file
• Sifter file modified to generate a "species" field that it then filters the samples by. This field is then discarded

[evergreen700]

BindingDB:

• Download.yaml fixed by removing name inconsistencies
• tsvParser.py created to handle the repeating columns at the end of the file
• transformTsv.yaml modified to read protien id's as a list instead of a single string, allows multiple protiens to be connected to a binding record
• protein_compound_association modified to add protein_structure property.

[evergreen700]

Made the requested changes

[evergreen700]

PharmacoDB parse is on now!