Hi,
Recently, I collected a dataset, but after data processing, I was only able to achieve a resolution of ~6 Å. I did not observe any high-resolution information in the reconstruction. What could be the possible causes of this? Could there have been an issue during data collection? The CTF estimation results look normal to me, so I’m wondering whether there are other factors I should consider.
Thanks,
Wendy
Did the 2D classes align well ? clear secondary structure features, etc ?
How did manual micrograph curation go (keep/reject%) ? I see the CTF histogram is well under 5A, but it does not look like patch CTF.
If those 2 tests pass, then this could be a box size / particle extract issue, too much F-crop ?
Agree with Mark; it’s really hard to tell just from a histogram and an FSC curve what the issue(s) might be. Although as FSC curve goes to ~1.8 Å, F-crop is unlikely to be the culprit.
If possible, share approximate molecular weight, 2D classes, total number of particles and particles in the classes you consider “good”. Also check some of the power spectra to see how strong the ice ring is. If you’ve seriously overpicked/not curated during inspect picks strictly enough, a lot of junk can still be hiding in the good classes, and this will hurt your results, so sometimes multiple rounds of 2D or multiple rounds of heterogeneous refinement will be necessary.
Sometimes it’s a more fundamental biochemical issue rather than an issue with acquisition or processing, though. For example, our lab has been struggling with one particular sample recently which, despite getting good CTF fits (<4Å for 99% of the data), the biochemistry has had serious problems resulting in the protein complex being severely damaged. After a lot of work with the collaborator, we’re finally getting good results (<2.5Å for consensus refinements)… might also be worth checking what your sample looks like via SDS and native PAGE and what FPLC traces look like.
