...since he was not using XDS... Oooops! My bad! I just recognized the phenomenon and assumed it to be from XDS. That explains the different column names.
I'll keep my eyes open next time. Cheers, Jose. ================================ Jose Antonio Cuesta-Seijo, PhD Carlsberg Laboratory Gamle Carlsberg Vej 10 DK-1799 Copenhagen V Denmark Tlf +45 3327 5332 Email josea.cuesta.se...@carlsberglab.dk ================================ -----Original Message----- From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Kay Diederichs Sent: Friday, July 17, 2015 11:40 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Negative CCanom Hi Jose, I don't think that Jacob's point was about XDS producing negative CC1/2-anom , since he was not using XDS. Concerning your observations: if there is zero anomalous signal then the CC1/2-anom is _not_ expected to be zero, due to noise in the data. Its absolute magnitude should then be expected to be low, though. Statistically, its sign will be positive in about 50% of the cases, and negative in about 50% of cases. XDS marks, in the tables in CORRECT.LP, with a '*' those values of CC1/2 and CC1/2-anom whose magnitude is statistically significant at the 0.1% level. For a case of zero signal (isomorphous or anomalous) and just the presence of random noise, you thus expect this star in one out of 1000 cases. Conversely, if there is no signal, but you do see significant (i.e. marked with '*') CC1/2 or CC1/2-anom values, then this (in 999 out of 1000 cases) indicates some systematic deviation from random noise. Could be radiation damage, as discussed in this thread, or some other effect, like also suggested in this thread (variable twin fraction). best, Kay
