Jerry, In similar situations, I've tried various solvent flattening approaches. Programs such as DM allow input of a manually created solvent mask. If you edit the mask created around your MR solution to include where you think additional protein density should be (using for example MAPMASK with the "border" and "combine" options), missing density can become stronger and not be washed out during solvent flattening.
Also, the newer versions of CNS can optimize a solvent mask that may avoid flattening out weak density. Personally, I would try DM using a wide range of different protein and solvent density levels along with a custom mask and look at the resulting maps (including NCS averaging) to see what's best. Good Luck! Paul Smith, Ph.D. --- On Wed, 7/22/09, Jerry McCully <[email protected]> wrote: > From: Jerry McCully <[email protected]> > Subject: [ccp4bb] model completion after MR with partial model > To: [email protected] > Date: Wednesday, July 22, 2009, 6:15 PM > > > > #yiv2140568612 .hmmessage P > { > margin:0px;padding:0px;} > #yiv2140568612 { > font-size:10pt;font-family:Verdana;} > > > > Dear ALL: > > Thanks a lot for the > input about EPMR statistics. After struggling for several > days, I finally got a promising solution from Phaser with a > partial poly-Ala model with 50% completeness. > > two molecules in one ASU: > solu set RFZ=4.2 TFZ=5.9 PAK=0 LLG=23 RFZ 3.8 > TFZ=19.0 PAK=3 LLG=110 LLG=123 > > After some initial refinement, some side chains have been > assigned with R-factor 50% and R-free 52%. In addition, > there are some extra density at two termini. > > We tried NCS averaging to improve the map but it did not > help that much. > > Now we are trying to build more residues in with 2.6A > data. > > Any suggestions will be highly appreciated. > > Jerry McCully > > > > Windows Live™ Hotmail®: Celebrate the moment > with your favorite sports pics. Check > it out. >
