Transferring annotations to geneproducts in between strains
Hi Rama, I have a quick question for you regarding how your group manually maps curated GO annotations from one bacterial strain to another of the same species. I was wondering if the best practice would be to use the same evidence code (e.g. IDA) for cases where I was looking at conservative ortholog calls for genes of the same species, different strains? Or should I use the ISO code and the with_from value (e.g. PseudoCAP:PA1777) to indicate the strain the function/process was We observed in? I used the former approach when I originally did some GO mapping to Pseudomonas aeruginosa strains in our database, but I'm now leaning toward the latter approach in order to make the logic behind the call more transparent. Do you have any opinion on which approach is best? Thanks for any help you can offer! Geoff
The most accurate representation for the bacterial strains would be to use ISS as the evidence because it is not an ortholog. Use the strain ID in the taxon column. We currently use K12 for the taxon for E. coli, so we don't make annotations at the species level for this. Should we allow annotation at the strain level? We think this should be allowed. Do we ever want to have strain information for mice? Eventually, we probably do, there are 17 strains to consider.