Hello,
I have two different simulations of a membrane protein in a E. coli model membrane, and I want to analyze the protein-lipids interactions, by generating 2D density maps. In each simulation, the protein moves/rotates different along the xy plane of the membrane, and ends up in a different orientation.
Is there a way to transform the coordinates of one simulation, ex. superposing one trajectory onto the other (taking the protein coordinates of the first simulation as reference) to have both in the same coordinates system and the proteins in the same orientation, so I can generate comparable 2D maps?
The attached pictured illustrates my question. The "protein vector" is only to show the protein orientation.
Thanks in advance.
Regards
Yasser
Hi Yassir, I like the way you visualize how proteins work. I do the same thing. Sometimes the experiments confirm what I see in my minds eye, most of the times they do not.
Membrane proteins are very tough to characterize because there are interacting with the lipid membrane in many different ways. You have integral membrane proteins, membrane associated proteins, lipid anchored proteins, lipid specific binding proteins, lipases, etc.
Have you tried flipping the model? That is having lipids as the core and having protein clusters binding to it.
I think you are looking for the Kabsch algorithm which will find the optimal translation and rotation for two sets of matched coordinates.
https://en.wikipedia.org/wiki/Kabsch_algorithm
Hi,
Thanks for your replies! My problem is not related with the algorithm itself. I can translate the system without problem. The problem is the boundaries artifacts of Sim2 when translated to Sim1.
Steingrimur Stefansson flipping the the model would means a completely different simulation, with a different rational in my case.
- Badges
- Science topic
- Similar topics
- Chemistry
- Biochemistry