On Friday 16 October 2009 15:49:16 Demet Arac-Ozkan wrote: > Dear All, > > I have a few questions about protein structure and fold. I will appreciate > your suggestions: > We have the crystal structure of a protein and it looks like it might be a > new domain. > > 1) Our structure looks like there are two domains however there are some > close interactions > between the two so we are not sure if we should define it as a single domain > or two domains. > I am looking for a program where I can enter a pdb file and get out domain > boundaries based > on the structure.
TLSMD analysis is a possibility. It will not give a definitive answer to the question "one domain or two?", but it may tell you where the boundary should be drawn if you do decide to treat it as two. http://skuld.bmsc.washington.edu/~tlsmd/ > For example something that can do a dot plot of the distances between > residues? > > 2) We are wondering whether distant homologues of this protein also have this > domain however sequence alignments are not reliable because there is little > sequence identity. For doing structure-based sequence alignments, which > programs can be used (other than SALIGN)? > > 3) What is the best way to decide whether a structure contains a new fold or > not? For example if DALI z scores are about 8 (half of the domain is similar > to one protein, the other half is similar to another protein), would you call > it a new fold? > > Thanks a lot in advance, > > Demet Arac > -- Ethan A Merritt Biomolecular Structure Center University of Washington, Seattle 98195-7742
