Question about new 3D refinement

[XUETING ZHOU]

Dear EMAN2 Community,

I'm trying to do the 3D refinement for a membrane protein, but the 3d refinement result is worse than the initial model. I tried several different strategies, but the result wasn't improved at all. 

For my initial model, I can see the clear membrane (image attached).

However, after I did a 3d refinement with a global search as following:

e2spt_refine_new.py --ptcls=sets/test_bin2_xf.lst --ref=sptsgd_03/output_cls1_ref.hdf --startres=30.0 --goldstandard --sym=c1 --iters=p,p,p,t --keep=0.9 --maxres=20.0 --parallel=thread:12 --threads=10 --loadali3d 

The density barely changed from referance and the orientations of the particles were completely messed up (image attached).

If I assigned an initial orientation for each particle(point out from membrane), and did 3d refinement with local search:

e2spt_refine_new.py --ptcls=sets/test_xf.lst --ref=sptsgd_03/output_cls1_ref.hdf --startres=30.0 --goldstandard --sym=c1 --iters=p,p,p,t --keep=0.9 --localrefine --parallel=thread:12 --threads=10 --loadali3d

the particles didn't align to the reference and the orientation didn't change at all (images attached).

I also tried to add a soft mask on 3d refinement, but didn't change the result.

Does anyone have any idea why this happen? Any suggestion/discussion will be greatly appreciated. Thanks in advance 😊

 

Many thanks,

Xueting Zhou

Baylor College of Medicine

[XUETING ZHOU]

Forget the mention, my EMAN2 version is 2.99.52.  Python version is 3.9.16. Thank you again.

[Muyuan Chen]

Hi, 

To make the initial orientation assignment work, you will need to orient the map so that the membrane is horizontal and the protein density (normal vector of the membrane) points outward, along the positive direction of the z axis. You may be able to do this by doing e2symsearch.py with a high c symmetry, or just rotate it with the xform processor in the Filtertool. Also remove the —localrefine option from the e2spt_refine_new command and add the —vector option.

Saying that, the fact that the global search does not work typically means there may not be enough signal from the particles in the first place, and the alignment is dominated by the membranes. Providing initial orientation may help but it can still be tricky. 

Muyuan

On May 31, 2023, at 11:09 PM, XUETING ZHOU <xtz...@g.ucla.edu> wrote:

Dear EMAN2 Community,

I'm trying to do the 3D refinement for a membrane protein, but the 3d refinement result is worse than the initial model. I tried several different strategies, but the result wasn't improved at all. 

For my initial model, I can see the clear membrane (image attached).

<ref.png>

However, after I did a 3d refinement with a global search as following:

e2spt_refine_new.py --ptcls=sets/test_bin2_xf.lst --ref=sptsgd_03/output_cls1_ref.hdf --startres=30.0 --goldstandard --sym=c1 --iters=p,p,p,t --keep=0.9 --maxres=20.0 --parallel=thread:12 --threads=10 --loadali3d 

The density barely changed from referance and the orientations of the particles were completely messed up (image attached).

<global search result.png><global search orientation.png>

If I assigned an initial orientation for each particle(point out from membrane), and did 3d refinement with local search:

e2spt_refine_new.py --ptcls=sets/test_xf.lst --ref=sptsgd_03/output_cls1_ref.hdf --startres=30.0 --goldstandard --sym=c1 --iters=p,p,p,t --keep=0.9 --localrefine --parallel=thread:12 --threads=10 --loadali3d

the particles didn't align to the reference and the orientation didn't change at all (images attached).

<local search orentation.png><local search result.png>

I also tried to add a soft mask on 3d refinement, but didn't change the result.

Does anyone have any idea why this happen? Any suggestion/discussion will be greatly appreciated. Thanks in advance 😊

 

Many thanks,

Xueting Zhou

Baylor College of Medicine

--
--
----------------------------------------------------------------------------------------------
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/eman2/4db4bbe6-56fe-4f61-be93-b4f352a3ff2fn%40googlegroups.com.
<global search result.png><global search orientation.png><local search orentation.png><ref.png><local search result.png>

[XUETING ZHOU]

Hi Muyuan,

Thank you very much for the prompt response. As you said, the alignment might be dominated by the membrane. I thought the membrane should be a strong feature, so I should be able to get at least a clear membrane structure for my 3D refinement output even though I might not see my protein. However, the membrane was not aligned well either.  Do you have any idea why this happens? Please correct me if my thought is completely incorrect. Thank you again.

Best,

Xueting Zhou

[Muyuan Chen]

I do not know enough about the system or your entire data processing procedure to make a good guess. I thought you showed membrane density in your average, which means the membrane actually got aligned. You still need to make sure the normal vector of the membrane points toward z direction, so you can visualize the orientation correctly. 

To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/6d006035-1f7d-4deb-9b0a-2cbe48fe4aa5n%40googlegroups.com.

[XUETING ZHOU]

Thank you very much! I'll try to rotate the map first to see how the thing is going.

Best,

Xueting

[XUETING ZHOU]

Just want to be clear,  the map that needs to be rotated is the initial model, right?

Best,

Xueting

[Muyuan Chen]

Any reference you use to start the refinement should be at the correct orientation. 

Muyuan

To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/13a397aa-f700-4ad3-bb4c-9f3c16de127cn%40googlegroups.com.