Hi Jess,

Pfam is a database of protein domain families (Google will take you there). Most proteins, especially in eukaryotes, consist of multiple domains. As you will know, these are parts of sequence that occur in (many) different proteins known (conserved to a certain degree, i.e., the sequences are similar but not necc. identical) and usually exhibit a conserved fold. Often, they also have a certain (partial) function within the protein as a whole, e.g. binding ATP. Pfam detects such domains using 'sequence profiles' of the respective domain families. In some cases, the result is not unequivocal, i.e., different families may 'fit' a certain region of the protein. While websites may then only report the 'best fitting' (highest scoring) set of domains, they may also decide to present all the information to the user and leave it to him/her which domains may be present. I hope this answers your question. A good way to compare the output of different sites is to go to Pfam directly, enter the UniProt ID of your sequence (or the sequence itself), and have a look at the domain structure reported!

HTH, Rob

Hi,

I would like to add that somebody (a scientist) has defined these domains at some point and added them to Pfam. Some entries will reference the source, i.e. the publication where this domain was first defined. However, as I recently found out myself, some entries may be outdated because maybe more sequences of related proteins from different species have now become available and the original assignment was biased. So be sure you trust this Pfam assignment! Another very useful server is InterPro.

BW,

Anja