I am trying to dock DNA to a protein (2 chains; it's a dimer), and to the best of my knowledge, there's not an out-of-the-box application that does this (there is another thread on here about that). However, I have used the RosettaScripts RosettaDNA design demo successfully, which reads, designs and scores a Protein-DNA interface.
I started with the demo in Protocol captures RosettaScripts/RosettaDNA; I was also able to remove the design aspect by removing the element, so now my final "designed" interface is the same sequence as the inputted pdb. And the simulation still retains local backbond movements because of the < DesignProteinBackboneAroundDNA> element.
However, the DNA does not translate or rotate, and I would like to add a mover to allow this. This is where I am stuck.
I've looked at RosettaScripts/ligand_docking and I see that I need to add a < MOVEMAP_BUILDERS > element and a new < MOVER > element (single and compund???). But I'm really confused as to how to go about this.
I know this question has been asked before, but I haven't seen any results or thoughts on how to implement this. Is it possible in Rosetta?
I've attached my current (and working - without adding rotation/translation movers) xml script.
The MOVEMAP_BUILDERS element is really for ligand binding - I'm not sure if it would be the best option. (Mainly because the ligand docking code tends to assume that what's being docked is a single residue.)
A better bet would likely be the protein-protein docking code. If you're lucky, it's written generally enough such that it will dock a DNA chain as well as a protein one. You may have to be careful about which options you use, as some may be protein-protein specific.
Baring that, you should be able to knock up something in PyRosetta. One possible complication is that Rosetta has limited facility to sample DNA flexibility. If you have DNA conformational change on binding, a Rosetta protocol might not be able to accommodate that without extensive additional protocols.