Micelle - particle subtraction

Tips
1

Hi,

Does anyone have a workflow for masking out the micelle?

I did the particle subtraction and I’d like to know how to proceed next.
Should I do the local refinement? In this case which are the inputs for this job? Should the mask be a negative of the micelle mask? And how about the input volume?

Thanks in advance!

Jessica

1 Like
2

Hi @fernandesj,

Have you already tried Non-uniform refinement? It’s specifically designed to optimally treat micelles in membrane proteins and does not require masking or subtraction, often achieving substantially improved map quality and resolution for protein domains. Check out the preprint here for results and details:

In our hands, non-uniform refinement generally works better than subtracting the micelle (since the micelle is actually present in every particle image, it’s just disordered).

2 Likes
3

Hi @apunjani,

Thank you for your reply!
Yes, I did try the NU-refinement and actually it works better for me.
However, I was trying to perform a 3D variability job and in the Tutorial Part 1 is written: " NB. For membrane proteins it is particularly helpful to mask out the micelle because otherwise, almost all the observable variability will be of the micelle and not the encapsulated molecule." How do I proceed in this situation?

Thanks in advance!

1 Like
4

Hi,

just a word to be precise, we are not talking about detergent micelles, but rather detergent belts around the membrane protein. The belt is a different object compared to the micelle, it doesn’t follow the same properties as the micelle in terms of size, number of monomers inside, and in case of detergent mixture they really do not behave the same at all.
An example is that there are 80 DDM in a micelle, while you can find 400 in the detergent belt around A 12TM membrane protein.

best
Vincent

2 Likes
5

Hi @vincent, thanks for the clarification - I was not aware of the structural difference between belts and micelles so this is very helpful. Thanks!

@fernandesj, for 3D Variability you should create a mask that includes (i.e. has value 1.0) the part of the structure you want to consider during 3DVA processing, and value 0.0 for parts to exclude (ideally with a soft falloff in between). You can create this in the simplest case by

  1. opening your current refined density map (the map, not map_sharp output) in Chimera, and noting the threshold at which only the desired part of the structure (maybe only protein, excluding the detergent) is visible.
  2. create a “volume tools” job in cryosparc and connect the refined map as as input. Set “type of output volume” to mask, and set “Threshold” to the threshold from Chimera. Then set the “Dilation radius” and “Soft padding width” to a distance in pixels that will define how much the selected density is dilated and softly padded to create the final mask. Once this job runs, you can download the mask and open in Chimera with your density map to verify that it includes the correct regions
  3. Connect the mask output from Volume Tools to your 3DVA job.
6 Likes
6

Thank you @apunjani!