martin2urban
commented
8 months ago PHI-base 4 is literature centric. This means if one gene is reported in two or more articles, than there could be many PHI-IDs per UniprotKB ID.
Open jseager7 opened 2 years ago
martin2urban
commented
8 months ago PHI-base 4 is literature centric. This means if one gene is reported in two or more articles, than there could be many PHI-IDs per UniprotKB ID.
The PHI-base curator guidelines state the following about PHI-base accession IDs in PHI-base 4:
My understanding is that there is a one-to-one mapping between a PHI-base accession ID (PHI ID) and the following triple of data types:
(PMID, Pathogen NCBI Taxonomy ID, UniProtKB accession number)
So, when grouping by the above data types, I would expect there to be one PHI ID for each group. Instead, there are over 100 groups where there is more than one PHI ID (a one-to-many mapping), as shown in the table below.
@martin2urban Is there some other criteria for assigning PHI IDs that I'm missing, or is the data below a result of an error in the logic used to assign the IDs?