*** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk ***
Hi Weikai, There are two kinds of merging going on here - summation of partially measured reflections (across two or more images) to give a full reflection intensity, and merging of symmetry related reflections which in theory should have the same intensity. When working with data from Mosflm, the partial reflections are not yet summed (scala does this later on) and the symmetry related reflections are recorded separately (scala merges them by default.) Pointless needs the symmetry related reflections to be recorded separately because it works by comparing the reflection intensities to derive the most likely symmetry axes, and from this determining the most likely pointgroup. Given that pointless works fine with data from XDS, I can only assume that the treatment of the partials is not important. Treating the data in P1 will mean that the "corrections" due to things like polarization are included, but without any assumptions of symmetry relationships being used, which could result in reflections being rejected. Hope this helps, Graeme ________________________________ From: [EMAIL PROTECTED] on behalf of [EMAIL PROTECTED] Sent: Tue 09/01/2007 2:31 AM To: [email protected] Subject: [ccp4bb]: Pointless HKL2000 and unmerged data *** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk <http://www.ccp4.ac.uk/> *** Hi All: I am trying POINTLESS for detecting real space group using data processed by HKL2000. Apparently, POINTLESS cries for unmerged data. I understand using Mosflm might be the best in this case, but I am not familiar with the program. So I did the following to rescue but not sure if it is the right thing. 1) Integrate (denzo) as P1 space group in HKL2000. 2) Scalepack in P1 with keyword "no merge". 3) Using COMBAT in CCP4 to covert to mtz and specify the input .sca file as "Scalepack file nomerged (no partials)". What I am more curious about, is what exactly does "unmerged" mean? Does it mean not reduced to asymmetric unit? Or something happened much earlier in data processing, like the "no partials" comment in COMBAT seems to suggest? If it is the latter case, why is it so important for symmetry analysis in POINTLESS? Regards, Weikai
