Hi Catherine, Remind me how exactly you are creating the axis in Chimera, and send me whatever script you are using to compute it yourself… —Eric

On Jul 18, 2017, at 8:24 AM, Catherine Jenifer Rajam Rajendran <catrajen@umail.iu.edu> wrote:

Hi Eric,

I am able to generate the axis. Like you said, we used the x,y,z axis subtracting it from centroid position and did the SVD. For most cases, the axis is same as Chimera but for few it is little different. I checked the number of atoms used between both and it is same. Are you using all the atoms or only specific atoms or is there any other thing I am missing?

Thanks, Catherine

On Fri, Jun 30, 2017 at 7:13 PM, Eric Pettersen <pett@cgl.ucsf.edu <mailto:pett@cgl.ucsf.edu>> wrote: Chimera simply calls numpy.linalg.svd (singular value decomposition) on an array of the xyz coordinate (after the centroid position has been subtracted from them). The documentation for numpy.linalg.svd is here:

https://docs.scipy.org/doc/numpy/reference/generated/numpy.linalg.svd.html <https://docs.scipy.org/doc/numpy/reference/generated/numpy.linalg.svd.html>

It looks like it in turn calls LAPACK routine_gesdd, which is documented, in general, here:

https://software.intel.com/en-us/mkl-developer-reference-c-singular-value-de... <https://software.intel.com/en-us/mkl-developer-reference-c-singular-value-de...>

It’s an implementation of singular value decomposition, described in general terms here:

https://en.wikipedia.org/wiki/Singular_value_decomposition#Relation_to_eigen... <https://en.wikipedia.org/wiki/Singular_value_decomposition#Relation_to_eigen...>

The math is pretty complicated.

—Eric

Eric Pettersen UCSF Computer Graphics Lab

...

On Jun 30, 2017, at 10:09 AM, Catherine Jenifer Rajam Rajendran <catrajen@umail.iu.edu <mailto:catrajen@umail.iu.edu>> wrote:

Hi Elaine,

Thank you for helping me understand it. Can you give me more details about how PDB file is decomposed to find the plane and normal. Like, how the eigenvectors are calculated from the atomic coordinates.

Thanks, Catherine

On Thu, Jun 29, 2017 at 7:23 PM, Elaine Meng <meng@cgl.ucsf.edu <mailto:meng@cgl.ucsf.edu>> wrote:

...

On Jun 29, 2017, at 2:14 PM, Catherine Jenifer Rajam Rajendran <catrajen@umail.iu.edu <mailto:catrajen@umail.iu.edu>> wrote:

Hi,

Hi Catherine,

...

I am trying to understand how the plane is defined in chimera. Can you please help me with that?

"Define a plane for the specified atoms. Eigenvectors/values are calculated from the atomic coordinates after subtracting the position of their non-mass-weighted centroid. The plane is anchored at the centroid and aligned with the first two eigenvectors (the third eigenvector is normal to the plane)."

Define a plane for the specified atoms. - specified atoms, i use 'sel protein’

OK, so if using the graphical tool (Axes/Planes/Centroids) it uses the selected atoms, meaning all the protein atoms if you had used “sel protein.” If using the “define” command, you can just specify “protein” directly in the command to use all the protein atoms, without making a selection.

...

Eigenvectors/values are calculated from the atomic coordinates - atomic coordinates, so x,y,z coordinates of all atoms. (Can you please explain a bit more like how and what you read from PDB file to define plane).

For the atoms you specified, there are ATOM lines in the PDB file with X,Y,Z coordinates. Each atom is a point in 3D (X,Y,Z).

...

after subtracting the position of their non-mass-weighted centroid - how is non-mass-weighted centroid is calculated

It is the centroid of the points (atomic positions), not sure how else to say it. The average position of all the atomic points is the centroid. Non-mass-weighted just means all the atomic points are treated equally, instead of heavier atoms being treated as more important than lighter atoms.

...

The plane is anchored at the centroid - which centroid is this? Is the protein moved to the origin(0,0,0) position?

The centroid is the same centroid as in your previous question. Nothing is moved.

...

Thanks, Catherine

You’re welcome. I hope this helps, Elaine ---------- Elaine C. Meng, Ph.D. UCSF Chimera(X) team Department of Pharmaceutical Chemistry University of California, San Francisco

_______________________________________________ Chimera-users mailing list: Chimera-users@cgl.ucsf.edu <mailto:Chimera-users@cgl.ucsf.edu> Manage subscription: http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users <http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users>