-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Hello Valerie-- > After much trial and error, the script is now outputting a complex of > my two proteins which are initially derived from input coordinates > (PDBs). In my initial calculations, I was using dummy gamma errors > (error in the PRE value) that were low. In these initial calculations, > the docked proteins reflected the PRE data. I have now started to > include the real errors based on my spectra, which are very large. > Now, the proteins are either not moving from their initial position, > or are coming close together but not in a way that reflects the PRE > restraints. Is there a way to weigh these errors? It may be that as a > result of significant T2 relaxation, some of the errors are inherently > large. Does anyone have experience in dealing with large errors in > gamma values from PRE data? Are you certain that you have the PRE parameters set correctly? You might look at the PRE section of one of the output .viols files to check this. Also, it sounds like you may need to increase the PRE force constant (using setScale). > > Also, from my understanding, PRE is also influenced by the population > of the transient complex. Is there a way to directly specify a low > population in the script so that the PRE restraints are weighted > differently (e.g., 2% v. 5% of the transient complex in existence)? > You can directly include species with alternate conformations. See e.g. Tang, C., Schwieters, C.D. & Clore, G.M. (2007) Open-to-closed transition in apo maltose-binding protein visualized by paramagnetic NMR. Nature 449, 1078-1082. However, I would first try to get the single-structure calculation working a bit better. best regards-- Charles -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) Comment: Processed by Mailcrypt 3.5.8+ <http://mailcrypt.sourceforge.net/> iEYEARECAAYFAkpTMRIACgkQPK2zrJwS/labmwCcD39TizVDPutFkDnp6Ap15J93 So4AnA43J6N1k204sYY53zKRsPAgQR4B =762v -----END PGP SIGNATURE-----
