Project

General

Profile

Actions

Bug #2658

closed

converting IMAGIC to mrc image using imagic_to_mrc() produces header issue for Relion

Added by Amber Herold almost 11 years ago. Updated over 10 years ago.

Status:
Won't Fix or Won't Do
Priority:
Low
Assignee:
-
Category:
Python scripting
Target version:
Start date:
02/12/2014
Due date:
% Done:

0%

Estimated time:
Affected Version:
Appion/Leginon 3.0.0
Show in known bugs:
No
Workaround:

Description

I'm trying to get Relion working inside Appion. Relion produces an error that it does not support VAX floating point conversion. It reaches this conclusion because ( amin > amax ) in the header of the stack file it is reading.

In Appion, the function doing the conversion is:

def imagic_to_mrc(imagic_name, mrc_name, yflip=False):

Relion runs fine doing the conversion outside of Appion with the following batch file:

#!/bin/csh -f
setenv IMAGIC_BATCH 1
###
### change these parameters
### 1) ../relion_recon49/start -- this needs to point to an IMAGIC stack (without the .hed extension)
### 2) ali -- this needs to point to the output .mrcs stack, without the extension
###

/opt/Imagic/stand/em2em.e <<EOF
IMAGIC
MRC
2D
MULTIPLE
../relion_recon49/start
ali
mrcs
YES
EOF

Anchi, does any of this ring a bell? Have an idea of what is going wrong?
If not, go ahead and assign it back to me and I'll try to track it down. I'm just not familiar with the tools for working with MRCs so I'll have a steep learning curve, and hope to get relion working quickly.

Actions #1

Updated by Amber Herold almost 11 years ago

Any chance the yflip implemented in #1906 requires us the swap the amin and amax values in the header?

Actions #2

Updated by Anchi Cheng almost 11 years ago

amin and amax in pyami imagic_to_mrc are copied form Imagic header. yflip_copy does not change that. If amin becomes smaller than amax, that would mean that imagic header location for amin and amax isn't what pyami/imagic.py assumed. That would be a place to start looking.

Actions #3

Updated by Anchi Cheng almost 11 years ago

  • Assignee changed from Anchi Cheng to Amber Herold
Actions #4

Updated by Amber Herold over 10 years ago

  • Status changed from Assigned to New
  • Assignee deleted (Amber Herold)
  • Priority changed from High to Low
Working around this by:
  1. Use stackIntoPicks.py to extract the particle locations from the selected stack.
  2. Run makestack2.py without ctf correction or normalization using the stackIntoPicks result as the Particles run.
  3. Run relion_preprocess with rescale and norm. Outputs .mrcs file.
Actions #5

Updated by Amber Herold over 10 years ago

  • Status changed from New to Won't Fix or Won't Do
Actions

Also available in: Atom PDF