Check species mappings genes/proteins

[DeniseSl22]

See WP2533_r107133, which has for three genes a Uniprot ID pointing towards an incorrect species (2x Arabidopsis thaliana and 1x Saccharomyces cerevisiae) compared to the species for the PW (Homo sapiens).

• Glycerol-3-phosphate 2-O-acyltransferase EC 2.3.1.198
• Phosphatidylinositol-4,5-bisphosphate 3-kinase EC 2.7.1.153
• Acylglycerone-phosphate reductase EC 1.1.1.101

In the RDF, mappings to Ensembl exist, which didn't start with "ENSG" (the code for Human IDs).
These mappings are not in the website linkouts, this uses the API and filters on species, where the RDF uses all mapping files simultaneously.
So we could use this additional mapping to our advantage, to track species which start with the wrong Ensembl code for the species of the PW.
I'll fix the three IDs above, but it would be good to make a Unit test out of this I believe.