This is a common task. Merely changing the header is not enough as your actual data is still indexed in the old spacegroup. You have to reindex the reflections and change the axis order to have a P21212 space group. There are several ways to do this but perhaps the easiest is to apply an appropriate transformation matrix using sftools REINDEX MATRIX keyword. You will be changing the axis order so don't forget to make the determinant negative. Don't forget to also change the header. Note that you can do the same in HKL or in a number of other programs.
Artem P.S. this exact question has come up before, see http://www.mail-archive.com/[EMAIL PROTECTED]/msg01240.html for an example. _____ From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of Jacob Wong Sent: Saturday, November 08, 2008 1:52 AM To: [email protected] Subject: [ccp4bb] CCP4i: P212121 to P22121 conversion Dear all, a dummy question: I had processed a data set with HKL2000 to produce a P212121.sca file. Phaser seems to find MR solution with P22121 "alternative" spacegroup option. Now I want to do rigid body refinement with Refmac in CCP4i, there is no longer option for me to choose P22121. Apparently I could have run back to the lab to reprocess the scalepack, but I would rather ask for your suggestions as a case study. I tried to modify the P212121 spacegroup label in the .sca file to P22121 (line 3 header), redo the scalepack2mtz, and then use the newer mtz file as an input for Refmac, result: Refmac complains not to have such a spacegroup. How do I get around with this, without a trip back to the HKL2000? Thank you! Jacob Wong
