UniProt Reference or the NCBI non-redundant version or ENSEMBL or RefSeq or IPI ...
•
written
by
Attila Csordas
I have to be honest, we are still using the last IPI_human. However, in the very near future we will move to UniProt. We have a policy of updating our databases every 6 months with regard to searching tandem mass spectra, since some groups require the latest sequences and others require the same version of database to search.
We tend to UniProt because it is the European-based database and because of the format of the annotation which we are used to. My understanding is that both NCBI nrdb and UniProt share information regarding proteins, so that proteins added to NCBI (generally those identified in the US) are quickly added to UniProt, and vice-versa.
As a slight aside, we also have a database which is based on IPI_human, but to which we have added some of the most common contaminants from other species - this allows us to gauge (in some limited fashion with humans) what contaminants are in the sample.
adding common contaminants is always a good idea, are those files publicly available somewhere or always customised?
Regarding common contaminants, the cRAP datbase is a good place to start: http://www.thegpm.org/crap/index.html
Log in to answer this question.
