Dear cryoSPARC developers,
I work with T=4 VLP which has a trimeric spike sitting on the icosahedral 3-fold axis. Prior to the local refinement, I preform symmetry expansion with symmetry “I” and generate appropriate mask around the spike. I then run local refinement in C1 using the mask and symmetry expanded set. But, since the spike is trimeric, is there a way to run local refinement in C3 symmetry?
Yes, there is an option to enforce symmetry in local refinement.
You will want to first use Volume Alignment Tools to recenter on the trimeric spike, and align the symmetry axis of the spike to Z.
Then you can run local refinement for the spike with C3 symmetry enforced, starting from the mask, volume and particle set output by Volume Alignment Tools.
See here for an example (local C4 refinement of AQP1).
That is an amazing paper! Great strategy! Thank you!
A couple of questions.
Do I have to specify center of mass and rotation angles for the spike in the Volume Alignment Tools, or the job will find these parameters itself?
If I have to provide new center and rotation angles to the Volume Alignment Tools, which program should I use to calculate them and how to do it?
Yep you you have to provide the center coordinates in volume alignment tools - you can do this by applying
measure center to the mask in UCSF Chimera.
It would be useful if there was an option in volume alignment tools to automatically recenter on the mask (this exists in local refinement for the purposes of specifying the fulcrum), but currently the center has to be specified manually, in voxels.
One other thought - if the VLP is large, you may want to re-extract the symmetry-expanded subparticles in a local box prior to local refinement. In this case, re-extract using the output of volume alignment tools with the desired box, remove duplicates (to remove subparticles that overlap - where one obscures the other), and reconstruct. Then you can proceed directly to local refinement with the particles and volume from the local reconstruction, after using volume tools to crop the recentered mask to the appropriate box size. This may or may not give better results, but processing will be a lot faster in a smaller box, particularly if non-uniform refinement is enabled.
Oli, thank you! I will try both approaches. If the smaller box will not give better results, at least I will learn how to migrate from the bigger to the smaller box.