[Chimera-users] Edge visualisation
goddard at sonic.net
Thu Aug 28 09:52:04 PDT 2014
You are right about the Measure and Colors Blobs tool — the “Show principle axis box” option shows a box aligned with the moment of inertia principle axes. I was mistaken in my previous email. The behavior is described in the documentation (almost every Chimera feature is described in great detail) — just press the Help button on the Measure and Color Blobs dialog to see:
"Show principal axes box [color well] - whether to show a bounding box in the specified color (default green) for each subsequently picked blob. A blob's bounding box is aligned with its principal axes of inertia, calculated assuming a constant mass per unit surface area.”
The Laplacian filtering is again described in the documentation. If you use the Volume Filter dialog press the Help button to see the documentation, or if you use the “vop laplacian” command you can see the documentation with command “help vop”. Anyways here’s what the docs say:
"Laplacian filtering. The Laplacian operation is a sum of second derivatives. Laplacian filtering is useful for edge detection but amplifies noise, so it may be necessary to perform smoothing such as Gaussian filtering beforehand. Finite differences v(i-1)-2*v(i)+v(i+1) along each axis are used, and voxels at the edge of the box are set to zero.”
You referred in your email to “second derivatives” and comparing them to zero. I thought you meant the six independent second derivatives that make up the Hessian. I guess meant the Laplacian (sum of second derivatives along x, along y, and along z.
When you run Laplacian or Gaussian filtering or any other map filter in Chimera it makes a new copy of the map. Chimera does not save maps in session files. So if you want that Laplacian map to appear in the session file you need to save the map to its own file before saving the session. You do this with the Volume Viewer dialog menu File / Save Map As…. This is a pain, and some day we will allow the option to save the maps in the session file (even though they can be very large). The session also does not save information about how to recalculate your Laplacian map, so you have to save that map separately for sessions to restore it.
On Aug 28, 2014, at 1:08 AM, Petr Brazda wrote:
> Hello Tom,
> Thank you for your answers and comments. I will try to work on this evaluation with my colleague, who is a skilled programmer. I think that measure and color blobs tool works differently (at least the boxes are not aligned with the data axes) - it finds the largest diameter and puts it as the "x axis", then finds the largest perpendicular diameter (y axis). I got two more questions:
> 1) how exactly is the laplacian calculated? I found this
> https://www.cgl.ucsf.edu/chimera/docs/ContributedSoftware/volumeviewer/gaussian.html. If I understand it well it says that it calculates only non-mixed differences along x, y, and z axes (no diagonal direction, no dxdy...) and you write that the second derivative is in fact Hessian. Could you please precise this?
> 2) I wanted to save a session where I have overlapped density and its lapacian filter and I had to save this filter to be able to save the session. However, when I reopened the lapla. filtered data, there was only one bar in the histogram in the surface style, which makes the map unusable for me so now I have to do the filtering and subregion setting each time I begin with the analysis. Is there a possibility to add another bar into the histogram?
> Thank you,
> Petr Brazda
> Institute of Physics of the ASCR, v.v.i.
> Cukrovarnicka 10/112
> 16200 Prague 6
> Czech Republic
More information about the Chimera-users