 |
 |
Le Mercredi 15 Juin 2016 20:48 CEST, "Edward A. Berry" <[log in to unmask]> a écrit: Just a remark: There is no need of calculating the Patterson function as the correlation coefficient of two origin-removed Patterson functions P1(r) and P2(r) is simply the correlation coefficient of the two sets of intensities I1(hkl) and I2(hkl). And removing the origin peaks is very important to avoid having high correlation with completely unrelated Patterson functions. Philippe D. > You can calculate the Patterson maps with ccp4 FFT program (specify keyword PATTERSON) and as you say you don't need phases or model. > > Then use Uppsala MAPMAN to calculate the correlation coefficient between two such maps. > > eab > > On 06/15/2016 10:28 AM, Rangana Sanjeewa wrote: > > Dear All, > > > > I would like to calculate the degree to which any given two Patterson functions are correlated to each other. Patterson functions are calculated from data only. Does anyone know any tool to do this? It is my understanding that Molecular Replacement programs do this, but they compare calculated vs. model provided. > > > > Thanks for your suggestions. > > > > Best, > > Rangana