CSAIL Event Calendar: Previous Series
|
Protein structure determination using model fragments from the Protein Data Bank Speaker: Ian Stokes-Rees , Harvard Relevant URL: The Protein Data Bank (PDB) contains over 70,000 macromolecular structures with atomic coordinates. Most of these models have been established using X-ray crystallography and a technique called Molecular Replacement (MR), used to bootstrap the model phasing data which is lost in the diffraction data. MR relies on a good selection of a model fragment, usually covering 15-40% of the unknown structure, that has sufficient structural homology for initial phase estimates to be calculated. Certain structures do not have known structural homologs, thus making MR difficult or impossible, and in other cases known (or anticipated) structural homologs do not provide successful phasing through the MR process. By harnessing US national cyberinfrastructure we have demonstrated that searches for good MR template models can be expanded to the entire fragment database derived from the PDB, the Structural Classification of Proteins, consisting of over 100,000 domains. Such a search can consume over 50,000 hours of computing, but has been shown to identify models suitable for phasing where no other technique had previously succeeded. This can also overcome challenges of low sequence identity and small model fragments (<10% of the structure). (PNAS Vol 107, 2010
See other events that are part of Bioinformatics Seminar Series 2011/2012 |
