Dear all!

How do You think, is it possible to run an MD simulation of the protein with a fully rotatable bond (changing amide rigidity) in the middle of the reconstructed disordered region?

I have a multi-domain protein lacking three disordered regions in each domain. For example, X-ray captured the entire domain folding, but some part (~40 residues, forming loops or something separated and flexible) is absent. There is no templates for homology reconstruction, the number of possible conformations is also enormous. Despite the evidence, that my suggestion is unreal (because of some shaperons, post-translational changes), what if I could create a randomly folded loop/region and then start an MD with some point (tricky topology parameters of the bond), dividing the loop in two segments? This 'weak' bond should allow a free rotation of two ends/segments of the loop, bounded to the domain by other ends. In this way I want to preserve the length of the loop (this modified bond in the middle of the loop will permit additional rotation, but will prohibit an increase in distance between neighboring residues, forming this bond). I thought about 'partial' REMD technique, but the domain and the loop are forming a whole strict contour due to the psi/phi geometries...

P.S. Simple MD didn't help me - the loop took some conformation, which stood stable across the MD. Maybe this is Ok. However, then why wasn't it reflected with the crystal structure?

Thank You!

More Alex Rayevsky's questions See All
Similar questions and discussions