UniProt: a hub for protein information.
Bottom Line: We present a new website that has been designed using a user-experience design process.We have introduced an annotation score for all entries in UniProt to represent the relative amount of knowledge known about each protein.These scores will be helpful in identifying which proteins are the best characterized and most informative for comparative analysis.
Related in: MedlinePlus
Mentions: UniProt has developed two complementary rule-based systems to automatically annotate uncharacterized protein sequences of UniProtKB/TrEMBL. These are UniRule, in which rules are created as part of the process of expert curation of UniProtKB/Swiss-Prot, and SAAS, in which rules are derived automatically from UniProtKB/Swiss-Prot entries sharing common annotations and characteristics. Both UniRules and SAAS use the hierarchical InterPro classification of protein family and domain signatures (15) as a basis for protein classification and functional annotation. These rules share a common syntax that specifies annotations—including protein nomenclature, function and important residues—and necessary conditions, such as the requirement for conserved functional residues and motifs. InterPro integrates signatures from the HAMAP (16) and PIRSF (17) projects within the UniProt consortium. The creation of family signatures in HAMAP and PIRSF is tightly linked to the expert curation of literature characterized template entries in UniProtKB/Swiss-Prot, which allows highly specific functional annotation even within large and functionally diverse superfamilies. As an example the HAMAP signature MF_01864 see Figure 4, which encapsulates the information from only seven peer-reviewed publications covering four experimentally characterized proteins that serve as templates to annotate the function of the bacterial tRNA-2-methylthio-N(6)-dimethylallyladenosine synthase family to over 11 000 UniProtKB/TrEMBL records. The UniRule pipeline also leverages the manual curation of UniProtKB/Swiss-Prot for the continuous validation of rules: annotations are refreshed at each release of UniProtKB/TrEMBL, and the consistency of each rule evaluated by comparing the predicted annotations with those of the current version of UniProtKB/Swiss-Prot. Only those rules whose predictions perfectly match UniProtKB/Swiss-Prot are retained for the current production cycle.