Next Previous Contents

3. Preparing your protein structures

For a normal-mode-based domain analysis, only a single protein structure is required, which must be available in the Protein Data Bank (PDB) format. DomainFinder makes an effort to deal with all common variants of the PDB format, but a correct interpretation is guaranteed only for files that follow the PDB format definition exactly, for example files from the Protein Data Bank itself. The PDB file may contain non-protein molecules (DNA strands, small molecules, etc.), but DomainFinder will extract only the peptide chains.

For a comparison of two conformations, you must have two compatible conformations, i.e. two PDB files that contain the same residues of a protein. DomainFinder verifies that the two files contain the same number of chains, and for each chain the same number of residues. However, it does not insist that the residue types be equal (in order to allow a comparison of mutations), and it cannot verify that corresponding residues in the files actually describe the same residues in reality. Since many published protein structures are incomplete, i.e. some residues are missing, you must take care that the two protein conformations you have are compatible, if necessary by manually editing the PDB files to remove residues that are in one conformation but not in the other. DomainFinder provides a visual help to verify that your conformations are indeed compatible.


Next Previous Contents