Pointers for the use of EMAN2's automask routine

[digvij...@gmail.com]

Hi all,

I am writing to inquire about some pointers about the use of EMAN2's automasking routine:

E.g. for your subtomo/subtilt refinement tutorial (https://blake.bcm.edu/emanwiki/EMAN2/e2tomo#Initial_model_generation_.2810_-_60_min.29), you seem to recommend the use of automasking for initial model generation and subsequent subtomo and subtilt refinement. 

1. For what kind of particles, would you suggest that the automasking is better (for atleast first few rounds of refinement) than designing/generating a mask (manual mask), which can be nicely interactively designed using e2filtertool.py.
   
   Is it for particles with a structured central core? Because I would imagine, the flood-filling algorithm (used in automasking) would perhaps work best in such cases, as it does the flood-filling from the central core to outside? 
   
   
2. Would you recommend the use of manual masks for certain cases when generating the initial model using the sgd?
   
   
3. Any additional thoughts that can help us choose between automask vs manual mask, would be very useful.

Thanks and cheers,

Digvijay

[Muyuan Chen]

The automasking routine is the same for single particle or tomography, and it seems to generally work for single particle cases. I only manually design masks for in situ proteins where there is no clear boundary between the protein and its surrounding environment. Most membrane proteins on native membranes fall into this category. 

It is better to not use masking for initial model generation. There are rare exceptions, and it might be safer to uncheck the Fourier option in those cases. 

--
--
----------------------------------------------------------------------------------------------
You received this message because you are subscribed to the Google
Groups "EMAN2" group.
To post to this group, send email to em...@googlegroups.com
To unsubscribe from this group, send email to eman2+un...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/eman2

---
You received this message because you are subscribed to the Google Groups "EMAN2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to eman2+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/e1b09ffd-2ec5-4723-aa38-05a168ca90c6n%40googlegroups.com.

[digvij...@gmail.com]

Thank you!

[Mike Grunst]

Dear EMAN2 community,

Following up on this, I am dealing with a membrane-bound protein. The initial model generation seems to look fine, but there is a very strong membrane density connected to the main structure. I had refined this structure to ~11A resolution in a different software and found that masking out the membrane was essential for good refinement.

I am having some trouble creating a custom mask with the e2filtertool which would exclude the membrane density. Could you help me to identify which mask options within the filter tool would help to draw/manually exclude this? I would then use this mask in the e2spt_refine_new.py command.

Alternatively, I could try and import a mask from IMOD (made with the imodmop + pad and taper options for a soft edge) or relion_mask_create using the initial model generated from EMAN2. Any general advice for using a mask created in another software?

e2version.py output:

EMAN 2.99 ( GITHUB: 2022-03-02 11:10 - commit: 983107a09 )
Your EMAN2 is running on: Linux-5.15.0-91-generic-x86_64-with-glibc2.35 5.15.0-91-generic
Your Python version is: 3.9.10

Thank you!

-Mikey

[Ludtke, Steven J.]

Sorry, I intended to respond to this when I let it through moderation (note that you aren't moderated any more, that's just the first post to avoid spammers).

Masks in EMAN2 for most purposes can be generic greyscale images ranging from 0 to 1. They can be sharp or soft. If importing a mask from other software you will want A/pix set correctly in the header and the box size in pixels to match. Both of these things can be checked with e2iminfo.py or the e2display.py browser, and modified if necessary with e2proc3d.py with the --apix and --clip options. I would suggest writing the mask as a .hdf file for EMAN2 use, though normally it should be fine if you leave it in some other format as long as Apix and nx/ny/nz are correct (and the alignment is correct, of course). 

It's hard to tell you exactly what to do in e2filtertool to get the result you're after if you're doing it by manipulating the map. In many cases a filter.lowpass.gauss -> threshold.binary -> etc  sequence can produce very nice masks with appropriate parameters, but it depends on the map. Certainly for some maps you'd need to combine that with some sort of geometric mask to get the desired results...  Often for complex masks, using chimera with segger, then selecting appropriate regions, followed by a low-pass filter in EMAN2 will produce a nice solution.

---

Steven Ludtke, Ph.D. <slu...@bcm.edu>                      Baylor College of Medicine
Charles C. Bell Jr., Professor of Structural Biology        Dept. of Biochemistry 
Deputy Director, Advanced Technology Cores                  and Molecular Pharmacology
Academic Director, CryoEM Core
Co-Director CIBR Center

To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/8a107c89-d8db-41b7-b317-fddbb8ca866cn%40googlegroups.com.