Location-conditioned statistics estimation methods

To illustrate the power of PISA wheels and dipolar waves to determine the structure of helical peptides and proteins in uniaxiaUy oriented lipid bilayers. Fig. 6a-c show SIMPSON/SIMMOL-simulated PISEMA spectra of an ideal 18-residue a-helix with a tilt angle of 10°-30° relative to Bq. In these simulations, we have tried to mimic experimental conditions by including a random distribution of the principal components of the chemical shift tensor and the dipolar coupling. The chemical shift distribution is 6 ppm for each principal element and has been established as follows we obtained — 85000 N isotropic chemical shifts reported to the BioMagResBank and selected only the — 31000 located in helical secondary stractures to have a data set independent on secondary chemical shifts. The standard deviation on the N chemical shifts for these resonances was — 6 ppm. With the lack of other statistically reliable experimental methods to establish such results for the individual principal elements of the N CSA tensor, we assumed the above variation of 6 ppm for all three principal elements. The variation of the H- N dipolar coupling was estimated by investigating the structures for a small number of a-helical membrane proteins for which the structures were established by liquid-state NMR spectroscopy. These showed standard deviations... 

---