In my experience, even when rather magically Phaser works with 20-25% identity, with 2.0 data you cannot always proceed to change and refine your structure. I would focus on the Se synchrtron data, or collect a very redundant set at the home source, which should allow you to find the Se and do some good phasing.
Tassos ________________________________________ From: CCP4 bulletin board [[email protected]] On Behalf Of Narayanan Ramasubbu [[email protected]] Sent: Thursday, December 10, 2009 4:40 PM To: [email protected] Subject: [ccp4bb] Difficult MR structures Deal All: I have a 2.0 A data for a SeMet protein (native crystal not available yet!) that has 6 Se sites. The cell comes out to be 65 67 101 and the angles are all very close to 90. The data set was collected in house with Cu 1.5418 A We integrated and scale in orthorhombic and the statistics are reasonable. There are 4 homologous structures and all of them have a sequence identity of 15-16 %. I have tried Phaser, AMoRe, EPMR with varying templates with polyala, polyser and varying the resolutions. I have also done a modeller and found best match with one of the four structures. I have used this derived model as well for input. I would like to know whether there other options (I am working on getting a good synchrotron data set at the peak wavelength for Se). Thanks Subbu PS: I am also trying P1 just in case.
