segclasslayer batch mode problem

[je...@crystal.harvard.edu]

Hi Carsten,

if I run segclasslayer with this input, and then manually select layer lines with the mouse in the gui, the generated plots looks like excpect (attached image1.png for layer line at 0.00417, image2.png for layer line at 0.00833).

Class average stack                      = proj1.hdf

Batch mode                               = False

Diagnostic plot                          = segclasslayer_diag.pdf

Class format                             = real

Pixel size in Angstrom                   = 3.205128

Precise helix width in Angstrom          = 580

Power spectrum resolution cutoff in 1/Angstrom = 0.04

Class number range to be analyzed        = (0, 0)

Layer line positions                     = 0.0,0.004166666

Pad option                               = True

However, in batch mode with "Batch mode = True" and:

1) Layer line positions = 0.004166666

segclasslayer crashes (attached log in segclasslayer.log)

2) Layer line positions = 0.0, 0.004166666

segclasslayer runs, but in the output plot (attached image3.png), the 0.004166666 profile is at the wrong postion (0. layer line instead of 1. layer line) and misslabled

3) Layer line positions = 0.004166666, 0.0083333

the curves in the output plot (attached image4.png) looks identical as in 2)

4) Layer line positions = 0.004166666, 0.0083333, 0.0166666

I cannot find the correct curve for the 0.00417 layer line in the plot anywhere (attached image5.png).

It seems to me that there is some bug if segclasslayer is run in batch mode?

All the best, Simon

Simon Jenni

Department of Biological Chemsitry and Molecular Pharmacology

Harvard Medical School 

[Carsten Sachse]

Dear Simon,

This is indeed a bug in the retrieval of layer line positions very close to the FT origin when running in batch mode. On the other hand you are trying to plot a 1/~250 Å layer line. This will be very close to the origin of your Transform. Do you really need to know this? Is this informative for you?

As an alternative, I would use the GUI mode where you can interactively retrieve what you are looking for. It looks like you found this part already.

Best wishes,

Carsten

[je...@crystal.harvard.edu]

Hi Carsten, thanks for your reply. I just started this project and I have to read more on helical diffraction theory.

Best, Simon