What open resources are the best for annotating genes as being secreted for mammalian genomes (e.g. rat, mouse, human)?
I know EnsEMBL proteins have signalP and TMHMM features which can be queried via EnsMART or SQL, and can query based on GO terms. Uniprot also has annotations for well known genes. What other resources exist that are reliable and easily accessed?
4 answers
The Secreted protein Database can be downloaded.
NCBI gene?
Have you tried the methodology used in this recent paper on a prediction method for secretory / non-secretory proteins ? They extracted positive dataset from Uniprot using sub-cellular localization annotations in the comments block.
Also, this paper (in which I was involved) described a pipeline for defining the set of secreted proteins. I think it's also available as a downloadable file...
Log in to answer this question.