Feature #2380: How to get best CTF estimation from database - Appion - Electron Microscopy Group

Actions

Copy link

Feature #2380

closed

How to get best CTF estimation from database

Added by Neil Voss over 11 years ago. Updated over 10 years ago.

Status:

Closed

Priority:

Normal

Assignee:

Amber Herold

Category:

Python scripting

Target version:

Appion/Leginon 3.0.0

Start date:

06/04/2013

Due date:

% Done:

Estimated time:

Description

Hi All,

I am sure many of noticed over the last year my complete tear down of the CTF infrastructure. Hopefully it is working for you now, if not please file a bug and let me know.

Anyway, the next step is when doing a CTF correction or displaying on the web the program will query all the CTF estimations for an image and try to decide the best estimation for that image. Currently, it is using legacy code, i.e., the one with the highest confidence value and it does not care which program it came from.

I propose that it uses my new CTF resolution criteria to get the best CTF estimation by default. In makestack, I think it should be a choice, but there are many calls for the best CTF in the code that you may not realize and adding an option to even program would be hard to implement on a short notice, e.g., FREALIGN recon. I have already made the changes to my code, but I have not submitted them to the truck. I wanted to hear from you first.

Thanks,
Neil

Files

Screen_Shot_2013-11-06_at_4.16.34_PM.png (61 KB) Screen_Shot_2013-11-06_at_4.16.34_PM.png

New makestack CTF settings

Neil Voss, 11/06/2013 02:16 PM

Related issues 4 (1 open — 3 closed)

Actions

Copy link

Updated by Scott Stagg over 11 years ago

Sorry for the slow response. I am in favor of adding your new criterion as an option, but I really want it to be completely backwards compatible. In other words, I would like the option to get the best CTF the old way as well as the new way. Is that possible?

Actions

Copy link

Updated by Neil Voss over 11 years ago

Hi Scott,

So far, I have only updated the web viewer part, I think it might be a little clunky to choose the type of sorting, but it could be added.

With makestack, the most critical part, it will be a drop down menu to select the type of criteria to choose. Other places that need an option is frealign and relion.

Actions

Copy link

Updated by Scott Stagg over 11 years ago

Well, I don't use the typical Appion infrastructure for generating Frealign parameter files, so it wouldn't be to big of a deal for me as long as you don't remove the old functions and instead make new functions with new names. I don't even have Relion installed, so that's not a big deal for me either.

Actions

Copy link

Updated by Amber Herold over 11 years ago

Hey Neil,
I ran Ace2, but it is failing because getBestCtfValueForImage was replaced with getBestCtfByResolution and the resolution method does not take the "method" parameter that is being sent to it.
Not sure if you forgot to check in a file with corresponding changes, or if the method param can just be removed from all the calls.

Also, is Ace2 in a good usable state? or are you planning more changes to it for the near future?
Thanks,
amber

Actions

Copy link

Updated by Neil Voss over 11 years ago

Hi Amber,

Sorry about that. The only reason that function is called is to print the current best CTF values. It is not used by the program.

I've made changes to ACE2, but I am sure if they are improvements, so I am not committing them at this time. Also, I wanted it to be stable for you.

Actions

Copy link

Updated by Anchi Cheng over 11 years ago

r17815 fixed the null first in sorting error Bridget found today in myamiweb display. At the same time, I changed the sorting to have two levels with 80_percent_resolution as the primary sort.

Neil, please see if this is reasonable. If so, it may make sense o do this at python level.

Actions

Copy link