[Peng Liu]

Dear all, 

I have got a few questions during tomogram reconstruction and it would be nice to get your feedback: 

1. When we reconstruct tomograms with Warp, does Warp correct CTF for the tomogram when doing back projection? I could not find this description. If Warp does not correct CTF for the tomogram, what would be the best way to correct the CTF before reconstruction if I would like to incorporate Warp in the pipeline? Warp estimates CTF during pre-processing, but I understand it is different and not for the tomogram reconstruction. 

2. I also reconstructed tomograms with Etomo. Is there an easy way to invert the contrast for the entire tomogram? In Warp, we have this choice, but I did not find this in imod. 

3. After writing stacks with Warp, there will be .st file as well in the imod folder. However, the header of this st. file generated by Warp is empty, and the tilt.angle and so on for data collection are all 0. In this case, I  let Etomo find the information from tilt file or enter the tiltseries information. However, I wonder this empty header issue is something general or not. 

Thanks in advance.

 

Best regards, 

Peng

--

Dr. Peng Liu

Max-Planck-Institut für biophysikalische Chemie
Am Faßberg 11, 37077 Göttingen
Germany

 : +49(0)17635209768
Email: peng...@mpibpc.mpg.de

[Alister Burt]

Hi Peng,

Responses inline below!

1. When we reconstruct tomograms with Warp, does Warp correct CTF for the tomogram when doing back projection? I could not find this description. If Warp does not correct CTF for the tomogram, what would be the best way to correct the CTF before reconstruction if I would like to incorporate Warp in the pipeline? Warp estimates CTF during pre-processing, but I understand it is different and not for the tomogram reconstruction. 

Yes, tomogram reconstruction in Warp is a set of tiled 3D reconstructions in Fourier space with the correct CTF for each image contributing to the reconstruction - this is described in the Warp paper if you wanted more details

2. I also reconstructed tomograms with Etomo. Is there an easy way to invert the contrast for the entire tomogram? In Warp, we have this choice, but I did not find this in imod. 

There might be some way with any number of CLI image processing programs but I don’t know off the top of my head - I would open in Python using mrcfile and multiply the volume by -1…

You can also invert contrast in 3dmod just for visualization

3. After writing stacks with Warp, there will be .st file as well in the imod folder. However, the header of this st. file generated by Warp is empty, and the tilt.angle and so on for data collection are all 0. In this case, I  let Etomo find the information from tilt file or enter the tiltseries information. However, I wonder this empty header issue is something general or not. 

We don’t try to perfectly recreate any particular metadata structure you may be expecting in the extended header - we’re just writing out a minimal set of information for a tilt series to be aligned by whichever external program people would like to use

Cheers,

Alister

Sent from mobile - apologies for brevity

On Apr 10, 2025, at 06:49, Peng Liu <penggr...@gmail.com> wrote:



--
You received this message because you are subscribed to the Google Groups "Warp" group.
To unsubscribe from this group and stop receiving emails from it, send an email to warp-em+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/warp-em/CALVQGL_cJ7Aft1UWLUFXTogJQo1Rv3Zf3RXL6pv9oHAXZS%3DgNw%40mail.gmail.com.

[Mathew McLaren]

Hi Peng,

If you want to invert the contrast for a tomogram generated by IMOD (or any MRC file), you can use the newstack command that is supplied with IMOD itself. Use the header command and check the maximum and minimum contrast values. For a tomogram with maximum of 1.5 and minimum of 0.5, you can then run:

newstack -sc 1.5,0.5 input.mrc output.mrc

The -sc option will scale according to desired maximum and minimum contrast values, and reversing the existing order will flip the contrast.

Cheers,

Mat

To view this discussion visit https://groups.google.com/d/msgid/warp-em/EB3AB7DF-832B-40AA-BBBE-79A3B41CB679%40gmail.com.

[Reika Watanabe]

Hi all,

In Warp, we have a function to invert the contrast during tomogram reconstruction. WarpTools, I could not find it. Do we have the same option while reconstruction in WarpTools?

Thank you very much!

With best wishes,

Reika 

[Huy Bui]

Dear Reika,

WarpTools ts_reconstruct \
  --settings warp_tiltseries.settings
  --angpix 2.12 \
  --dont_invert

You can choose --dont_invert to not inverting contrast (protein black) or use it to get protein white contrast.
This has a totally different meaning to --dont_invert during earlier import step.

H

To view this discussion visit https://groups.google.com/d/msgid/warp-em/06b748a0-fb16-4169-a603-1ec0192e9538n%40googlegroups.com.

[Reika Watanabe]

Dear Daniel and Hui,

It worked beautifully! 

Hui, thank you very much for clarification about these --dont_invert options at different steps have very different meanings.

In Warp, it was indeed clear, invert tilt angles versus invert contrast. 

Thank you very much!!!

Reika 

To view this discussion visit https://groups.google.com/d/msgid/warp-em/CADBiWEa4WuKCLLJyV-Jku732vpmKiJ01GJUWRppLfkOWUH_SEg%40mail.gmail.com.

[Alister Burt]

Hi all,

FWIW: I think (but haven’t checked) that inverting contrast compared to the default white on black output breaks template matching downstream and wouldn’t recommend it if you’re doing that.

Most image viewers have ways to invert colormaps built in, in 3dmod it’s F11… should really add this to the docs as it’s somewhere where Warp is different to other packages

Cheers,

Alister

Sent from mobile - apologies for brevity

On Sep 24, 2025, at 01:54, Reika Watanabe <reika.wa...@gmail.com> wrote:



To view this discussion visit https://groups.google.com/d/msgid/warp-em/CAM1UCztbcDG2dNa6L5Ehn2H7v-APX6gqv-fPvOHBOinL2za-%2Bw%40mail.gmail.com.

[Huy Bui]

Hi Alister,

I used the default contrast and have no problem with template matching with ribosomes. I actually never tried with don_invert contrast for Warp template matching.

Best,

Huy

[Hamidreza Rahmani]

I have used dont_invert for pyTOM and seems like results are similar either way (as long as the template is also inverted). But i have not checked it carefully yet.

Best,

Hamid

To view this discussion visit https://groups.google.com/d/msgid/warp-em/CADBiWEYfAncg8B2uLhDrqzTA57rXq4XsO2TXs9V-jQj18vSWCA%40mail.gmail.com.

[Daniel Cai]

I’ve also noticed that template matching in gapstop with inverted/noninverted tomograms with corresponding contrast templates results in near-identical template matched peaks. I have not tested if template matching functions when the contrast of the template and the tomogram do not match (im assuming it will be problematic).

Daniel

To view this discussion visit https://groups.google.com/d/msgid/warp-em/CAA59SzMz305Mx-a4VZsddMZ1iuuAFEgx5_F8j%2B0JJuKVk189UA%40mail.gmail.com.

[Daniel Cai]

Hi Reika,

Try the:

invert/no-invert flag in ts_reconstruct.

Daniel

Get Outlook for iOS

---

From: war...@googlegroups.com <war...@googlegroups.com> on behalf of Reika Watanabe <reika.wa...@gmail.com>
Sent: Tuesday, September 23, 2025 8:21:40 PM
To: Warp <war...@googlegroups.com>
Subject: Re:

 

To view this discussion visit https://groups.google.com/d/msgid/warp-em/06b748a0-fb16-4169-a603-1ec0192e9538n%40googlegroups.com.