Hi everyone,
During heterogeneous refinement, my model shows extra density that doesn’t exist in 2D classification, and theoretically, it’s impossible for the model to contain the extra density. Is there something wrong with data processing? We are using cryosparc version v4.6.2.
All the best,
Yan
Would it be possible that there is something wrong during particle alignment, resulting in extra density that appears after heterogeneous refinement?
Hi Yan,
This is hard to diagnose without some idea of what the 2D classes and heterogeneous refinement look like. What resolution are you at and are the features you see in the map consistent with that resolution?
het refine requires volume inputs as reference and they don’t change drastically during maximization step relative to other job types. double check the volumes you input. It’s conceivable that you don’t know your 2D quite so well - what you want to find out is if the het refine has identified an unexpected part of the protein/complex, or if the job is not performing well. Yes, sometimes it happens that small densities on the periphery of (and separated from) your protein/complex can persist and make the masking/classification go a little haywire. Check that the volumes you input looked normal at multiple thresholds. Maybe, just maybe, you will identify something unexpected and important. Agreed it would be easier to characterize with 2D classes and 3D volumes or slices if not private.