Using CTF corrected particles in NU-refinement (NEW) without further CTF refinement

3D Refinement
1

Hi,

Will Non-Uniform Refinement (NEW) use previously CTF-refined particles even if I don’t turn on CTF refinement options? Or do I have to use Non-Uniform Refinement (Legacy) and turn on "Enable higher-order CTF)?

After running 1) Local CTF Refinement and 2) Global CTF Refinement (Tilt/Trefoil/Spherical Aberration/Tetrafoil and Anisotropic Mag), plots show good sign of global CTF fit (noise in residual plot).

When I proceed to 3) NU-refinement (NEW) w/ local and global CTF refinements (Tilt/Trefoil/Spherical Aberration/Tetrafoil and Anisotropic Mag), plots show the fit getting worse. Also, the new spherical aberration was 2.8 at step 2 and shot up to 3.2 during step 3.

Because Non-uniform Refinement (NEW) does not have ‘Enable higher-order CTF’, I wasn’t sure that it will automatically use previously CTF-refined particles.

Thank you,
Joon

2

Yes, it will use previously refined CTF params, you don’t have to refine them on the fly.

1 Like
3

Thank you so much for confirming this!!

Just out of curiosity, if I run multiple global CTF refinements testing different parameters, will the particles have different, unique CTF values from each other? Or will they have the particles from the job that finishes last? I’m asking this because at the end of CTF refinement, it says it updates the value for the group.

Thank you in advance.

4

The particles from each job will have the CTF parameters refined in that job. They are independent jobs, the order the jobs are run in doesn’t matter unless the output of one is connected to the input of another

1 Like
5

Awesome! Thank you for the clarification!