Stephanie DeLuca - RosettaEPR: An integrated tool for protein structure determination from sparse EPR data

Document created by Stephanie DeLuca on Aug 22, 2014
Version 1Show Document
  • View in full screen mode

  Publication Details (including relevant citation   information):

  Stephanie J.   Hirst   , Nathan   Alexander,  Hassane S.   Mchaourab,  Jens   Meiler

  S.J. Hirst et   al. / Journal of Structural Biology 173 (2011) 506–514


  Site-directed spin labeling electron paramagnetic resonance   (SDSL-EPR) is often used for the structural characterization of   proteins that elude other techniques, such as X-ray   crystallography and nuclear magnetic resonance (NMR). However,   high-resolution structures are difficult to obtain due to   uncertainty in the spin label location and sparseness of   experimental data. Here, we introduce RosettaEPR, which has been   designed to improve de novo  high-resolution protein structure prediction using sparse   SDSL-EPR distance data. The “motion-on-a-cone” spin label model   is converted into a knowledge-based potential, which was   implemented as a scoring term in Rosetta. RosettaEPR increased   the fractions of correctly folded models (RMSDCα < 7.5 Å)   and models accurate at medium resolution (RMSDCα < 3.5 Å) by   25%. The correlation of score and model quality increased from   0.42 when using no restraints to 0.51 when using bounded   restraints and again to 0.62 when using RosettaEPR. This allowed   for the selection of accurate models by score. After full-atom   refinement, RosettaEPR yielded a 1.7 Å model of T4-lysozyme, thus   indicating that atomic detail models can be achieved by combining   sparse EPR data with Rosetta. While these results indicate   RosettaEPR’s potential utility in high-resolution protein   structure prediction, they are based on a single example. In   order to affirm the method’s general performance, it must be   tested on a larger and more versatile dataset of proteins.

  Address (URL):