Simple Modular Architecture Research Tool (SMART) and its accompanying database provide a web service to annotate proteins by searching 1200 manually curated Hidden Markov models (HMMs), which are based on over 100 million proteins from UniProt, Ensembl and STRING.
Furthermore, it can annotate proteins from genomic sequences; its database comprises STRING, itself a database containing predicted proteins from only complete genomes.
Pre-calculated analyses of proteins from all three databases are available. In general, the annotation includes not only domain prediction, but provides taxonomic and other analyses.
Importantly, SMART provides the option of using HMMER on the Pfam database of profile-HMMs, which provides differing statistical analyses and contains additional protein families, respectively.
Find out more in the following paper:
Letunic, I., Doerks, T. and Bork, P., 2015. SMART: recent updates, new developments and status in 2015. Nucleic acids research, 43(D1), pp.D257-D260.