Ewa,
What is the best and user-friendly program to calculate a surface of
the interface between two proteins?
If you want a list of interface residues...
The Protein-Protein Interface Database
(http://ppidb.cs.iastate.edu/ppidb/InterfaceResidues) is easy to use,
but it is only for structures in the PDB. It will identify interface
residues using a distance based measure or loss of surface area upon
binding.
You could use the Protein Structure Analysis Package
(http://iris.physics.iisc.ernet.in/psap/), but that server is down at
this moment.
Another option is to use the WHAT IF Server
(http://swift.cmbi.ru.nl/servers/html/index.html). Its atomic contacts
are based on the distance between VDW surfaces. Since it reports all of
the contacts, you'll need to look for contacts involving different chains.
The KFC Server (http://kfc.mitchell-lab.org) predicts binding hot spots
within a protein-protein interface. You could use this if you don't
mind defining an interface residue as any residue within 4 A of the
binding partner. The output only lists interface residues.
(Disclaimer: I was a principal developer)
If you want to calculate the buried surface area...
I would split the complex into separate files, then run MSMS on the
unbound structures and the bound structure. You can download the
program at http://www.scripps.edu/~sanner/html/msms_home.html, or use it
online at http://helixweb.nih.gov/structbio/basic.html. You'll need to
perform this calculation:
Buried_Surface_Area = ASA_bound - ASA_unbound1 - ASA_unbound2
ASA = Accessible Surface Area
Perhaps someone else knows of an easier way to do this.
Or two domains?
If you want to look between two domains, you'll probably want to break
the molecule into two chains, and then run the analysis. Since you're
talking about domains, there should probably be a place to make a clean
break. If so, it's just a matter of inserting a TER record.
Is there any common criteria how to define boundary for that interface
(other than van der Waals distance)?
As stated above, you can use loss of surface area upon binding (assumes
proteins are rigid bodies), distance between C-alpha atoms, distance
between residue atoms, or distance between VDW surfaces. They'll all
produce similar results, but to my knowledge, there isn't a preferred
method. It's a matter of personal preference.
From a programming point-of-view, the distance between C-alpha or
residue atoms is preferred because it's the easiest to implement and the
fastest to compute.
Regards,
Steve Darnell
--
Steven Darnell
Department of Biochemistry
University of Wisconsin-Madison
Madison, WI USA
