Dear Sjor,
I am currently on the particle polishing step of processing my protein (~600kDa) and have encounter the same error after multiple runs. I receive the following error:
+ Reconstructing two halves of shiny particles ...
== Reading input half-reconstructions:
+ half1-map: Refine3D/run04_movie_ct20_data_shiny_half1_class001_unfil.mrc
+ half2-map: Refine3D/run04_movie_ct20_data_shiny_half2_class001_unfil.mrc
== Using a user-provided mask ...
+ input mask: Refine3D/postprocess_run04_automask.mrc
== Masking input maps ...
+ randomize phases beyond: -247.678 Angstroms
Postprocessing::run ERROR: FSC curve never drops below randomize_fsc_at
I have looked up and tried a few suggestions from previous post regarding this error including the additional argument of --bfactor_running_avg 3 and it did not work. My bfactor.star file includes frames with values and frames with -nan for BFactorUsedForSharpening and FittedInterceptGuinierPlot columns. What is the best way to check what is causing this error or to resolve this issue? My input parameters for particle polishing includes: Running Average Window: 7 (same value used in the "Refinement with Movie-Particles" step 8.3 in tutorial), Stddev on Particle Distance (pix): 300, Highres-limit Per-Frame Maps (A): 6, and Lowres-limit B-factor Estimation (A): 20.
Also, I have noticed a few output discrepancy during the Particle Polishing Step. In my movies, I have 32 frames, but I extracted frames 3-27 at the "Extraction the Movie-Particles" step. I noticed that the output files includes frames 1-25; therefore, I was wondering if Relion always reads the first frame as 1 (if frame 1 is actually frame 3 from the movie)? If so, then I found it odd that the first frame did not have any cross-sectional image when displayed it RELION. I saw that there was a bug in the movie extraction and corrected it as described in known issue website (http://www2.mrc-lmb.cam.ac.uk/relion/index.php/Known_issue).
Thank You in Advance,
Kevin
|