DO NOE DISTANCES CONTAIN ENOUGH INFORMATION TO ASSESS THE RELATIVE POPULATIONS OF MULTI-CONFORMER STRUCTURES

The feasibility of determining the relative populations of multi-confo rmer structures from NOE-derived distances alone is assessed. Without cross-validation of the NOE restraints, any population ratio can be re fined to a similar quality of the fit. Complete cross-validation provi des a less biased measure of fit and allows the estimation of the corr ect population ratio when used in conjunction with very tight distance restraints. With the qualitative distance restraints most commonly us ed in NMR structure determination, cross-validation is unsuccessful in providing: the correct answer. Other experimental sources are therefo re needed to determine relative populations of multi-conformer structu res.