Symmetry operation

Hello team,
I wish to ask a question I am solving a structure of a complex consisting of the 6 same monomeric proteins having hexagonal symmetry interacting with two different proteins. So I have tried putting C6/C1 symmetry both? The problem I face is when I put C6 the other protein also tries to fit in that symmetry which is impossible owing to its unsymmetrical structure. Could you shed some light on what symmetry operation I can apply?

If it is asymmetric when interacting, you must treat it as asymmetric. You could potentially work in C6, symmetry expand and 3D classify, but that can prove tricky if the interactions are irregular.


Hi @BhawnaMishra! I agree with @rbs_sci’s comment — if you have an asymmetric binder, you will need to treat the whole particle as C1 symmetric. However, you can still take advantage of the symmetry in your particle in a few ways:

Symmetry relaxation

When we perform symmetry relaxation, we still perform alignments using C1 symmetry, but after we find the pick an alignment we check the other, symmetry related positions to see if they’re better. This will probably help you get a good consensus refinement of your whole particle, especially if your binder is typically only in one position per particle.

You can read more about symmetry relaxation in our guide article on the topic.

Symmetry expansion

Once you have a good consensus alignment, you can perform symmetry expansion to improve the map quality for your individual monomers. In symmetry expansion, we create an additional “copy” of the particle for each symmetry-related position. So in your C6 case, for each particle position, you’d have six copies of the particle, with each particle centered on a different monomer.

Now that you have a “separate” copy of each monomer, you can create a mask around a single monomer and then align and classify these individual subunits to improve the map of the monomers both with and without the binder.

It is very important that you do not perform global refinements of particles which have been symmetry expanded. This has the possibility of creating duplicate particles in separate half sets, which will create FSC artifacts. You should only perform jobs like 3D Classification, which do not alter particle poses, or Local Refinements, which only search a limited range of poses near the starting position.

I hope that helps!


Thankyou so much team. I shall try this method and surely update my results here.