Curious FSCs with highly symmetrical particles

I would greatly appreciate some help interpreting attached FSC curves. Both are from homogeneous refinements, 1st is of Apoferritin with O sym applied and other is of a sample with C14 sym applied. Interesting features are the spikes in FSCs at high resolution. These give the rise to the reported resolutions of 2.9 and 3.8. I have done a number of Apoferritin reconstructions before and all of them follow the standard gradual drop of without the spike. I have ran local resolution map for the ApoF collection and it reports voxels between 2 and 3 A in the protein density, but also in the noise around the particle.

Thank you!
Apo_FSC_Osym S1_FSC_C14sym

Hi Andrey - this looks a little like what one sees in the presence of substantial beam tilt. Have you tried beamtilt refinement in relion?

Cheers
Oli

With a high symmetry the artifacts in the first example can be caused by overlapping particles picked from dense micrographs.

CTF errors also can produce that kind of result. If I saw the first example I would definitely try at least defocus refinement.

Dear Oli and Daniel,

Thank you for your suggestions. Oli, I will feed them into Relion to check for beam tilt. Daniel, particles in the images for the 1st dataset are generally well separated. I will run per-particle ctf in Relion as well to check on defocus estimation.
Interestingly, FSC for the 2nd dataset “normalized” after non-uniform refinement.