Just to be clear, the CCP4 data processing programs (SCALA and its replacement AIMLESS) always give you I+ and I- in the output. The only difference between "anomalous on & off" is in the outlier rejection, since if you have a large anomalous signal you don't want to reject as "outliers" reflections with a good strong anomalous difference. AIMLESS now automatically detects whether there is a substantial anomalous signal and switches this option ON if there is (unless you specify the option explicitly). There are also different Rmeas etc values within I+/I- sets and over all data.
In the scaling, as James points out, it is nearly always best to ignore the I+/I- distinction, unless you really have a huge anomalous signal (almost impossible for macromolecules), since you want to try to minimise anomalous differences to reduce systematic errors, so that what is left is more likely to be real signal. SCALA allows you the (unrecommended) option to separate I+ and I- in scaling, but I haven't programmed this in AIMLESS since I have never seen a case where it would be a good idea. As far as I know, in CCP4 you only lose I+ and I- if you explicitly remove them. Phil On 13 Jun 2012, at 08:03, Murray, James W wrote: >> I think there is a misconception floating around that processing your >> data with "anomalous turned on" will somehow degrade the quality of >> "normal" intensity data. > > I can think of very few circumstances when I would NOT want anomalous data, > yet for many data processing pipelines, it is the default not to give you the > I+ and I- separately. Anomalous data are very useful for locating metal ions > that you might not even have suspected to exist in your structure. Can I make > a plea that all data processing packages/pipelines give you anomalous data by > default? Can anyone think of a good reason why they shouldn't? > > James > > -- > Dr. James W. Murray > David Phillips Research Fellow > Division of Molecular Biosciences > Imperial College, LONDON > Tel: +44 (0)20 759 48895 > ________________________________________ > From: CCP4 bulletin board [CCP4BB@JISCMAIL.AC.UK] on behalf of James Holton > [jmhol...@lbl.gov] > Sent: Wednesday, June 13, 2012 3:47 AM > To: CCP4BB@JISCMAIL.AC.UK > Subject: Re: [ccp4bb] Fun Question - Is multiple isomorphous replacement an > obsolete technique? > > I think there is a misconception floating around that processing your > data with "anomalous turned on" will somehow degrade the quality of > "normal" intensity data. I'm not exactly sure where this rumor comes > from, but I imagine it has something to do with confusion about all > the various "anomalous" options different scaling programs have. For > example, some programs offer the option to treat all I+ and all I- as > completely separate data sets, scaled and merged independently. I > think this is called "scale anomalous" in SCALEPACK and "intensities > anomalous" in SCALA. Neither of these is the default because such > treatment is only helpful if the anomalous signal is absolutely huge > (I have only seen this once). So, I imagine people who have never > done experimental phasing (there are lots of them!) might read things > like "Switching ANOMALOUS ON does affect the statistics and the > outlier rejection" in the SCALA manual and decide that they had better > turn off all those evil "anomalous" things. Then they tell their > students to do the same, etc. > > -James Holton > MAD Scientist > > > On Tue, Jun 12, 2012 at 3:17 AM, Eleanor Dodson > <eleanor.dod...@york.ac.uk> wrote: >> Why would anyone ignore the anomalous data they had collected? It will >> always help the phasing, and decide the hand for you.. >> Eleanor >> On 6 Jun 2012, at 03:55, Stefan Gajewski wrote: >> >>> Hey! >>> >>> I was just wondering, do you know of any recent (~10y) publication that >>> presented a structure solution solely based on MIR? Without the use of any >>> anomalous signal of some sort? >>> >>> When was the last time you saw a structure that was solved without the use >>> of anomalous signal or homology model? Is there a way to look up the answer >>> (e.g. filter settings in the RCSB) I am not aware of? >>> >>> Thanks, >>> S. >>> >>> (Disclaimer: I am aware that isomorpous data is a valuable source of >>> information)
