Starting from molecular replacement models at 3.0 A is very tough.
Unless the identity/similarity of the search model and expected model
is very high, re-building the model to represent your structure,
is a very tough job, even for experienced crystallographers at 3.0
I suspect that the reason your Rfree is not going down (which must be your main worry
right now) is that you are being too conservative during rebuilding.
Try to delete and add regions according to the density map and see how it goes.
A couple of tricks that are useful, is to do some density modification (dm, resolve, etc)
and/or run arp/warp (it will most likely get rid of most of your model at 3.0, but it
will likely produce a lower Rfree and better map that can help you build manually).
Tassos
On May 21, 2009, at 15:40, Sravanti Vaidya wrote:
Hello CCP4i GURUs,
I am working on a 3A resolution structure and I recently got a solution from
Phaser. I have just started building and refining the structure. I started
with Rfree-43 , R-41.7 and the first round of building and refining gave me
Rfree- 40 and Rfac- 35.5. The Rfree and R do not decrease to the same extent
during refinement (Refmac5). Further rounds of refinement increases the
difference between R and Rfree (Rfree-38 Rfac- 30)
I am worried that since this is just the beginning further rounds of
building and refinement might drastically increase the gap between R and
Rfree. On the positive note the FOM is increasing with refinement and the
maps look better. I also see new density at several regions where I can
build into.
I tried to change the weighting factor and geometric parameters after I read
the earlier posts here. The references noted earlier were extremely helpful.
Here are some other details of my structure which might help you
Space group- P1 21 1
Resolution- 3A
Solvent content- 39%
unit cell parameters- 63.3, 90.72 , 86.11, angles- 90, 91.5 , 90
Please enlighten me with your suggestions. Is it fine to go ahead with this
refinement or is there a way I can reduce this difference between R and Rfree??
Your help will be greatly appreciated. As this is my first structure I am
excited about it!!!
Thank you
Sravanti