I agree there are probably much better and simpler ways of doing this. But to answer the question, you could probably use MrBUMP to do this, using the local PDB file option. It will require a bit of scripting to deal with all the file names, and will take forever.
Other points: 1) you don't need to take every frame of the simulation 2) you probably need to minimise each frame 3) you could do clustering to get fewer diverse structures 4) you could do PCA, and use the lowest modes to distort the structure, a la normal mode methods But Poul's approach is still better ... Martyn ________________________________________ From: CCP4 bulletin board [CCP4BB@JISCMAIL.AC.UK] on behalf of Poul Nissen [p...@mb.au.dk] Sent: 30 March 2011 07:10 To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] MR with MD trajectories Dear Wandu, It might be smeared density because it is missing from your phasing model, so no reason to reinvent the wheel with MD and Phaser. You see electron density features for the domain and you know its basic structure. You can place the domain manually - it would take you five minutes and then you can proceed to refine the position by rigid body refinement and ordinary refinement/rebuilding Poul On 30/03/2011, at 05.43, Vandu Murugan wrote: > Dear all, > I have a protein molecule with two flexible domains. For a 2.8 angstrom > data, I could get MR solution for the larger domain but not for the smaller > domain (around 80 aminoacids). The second domain appears as extra density > showing its presence , but does not allow manual model building on it, since > it appers to be smeared. I would like to run a phaser with some 30000 > structures that I have for this protein from a simulation. Would it be > possible to do this in my computer? If so, how can I do this? Any > suggestions on this will be appreciated. Thanks in advance. > > regards, > Wandu > >
