[Chimera-users] Error opening larger proteins

[Eric Pettersen]

Hi Catherine,
	It turns out that the singular value decomposition function that Chimera uses to get the eigenvectors and eigenvalues for the plane computation uses a lot of memory if you use a large number of atoms to define the plane (or axis for that matter).  There is a non-default setting for that svd function that I just discovered that uses much less memory and seems to produce identical results, so I have switched Chimera to using that setting.  The switched version will be available as the next daily build.  Look for daily builds dated March 1 or newer.
	Also, you may be able to get your current version to work if you simply use less atoms to compute the plane.  The selection you were using — “sel :.A,.B,.C,.D” — will not only select all the protein but also all the waters, ligands, and ions.  Perhaps just “sel protein” would be more appropriate (depending on what you want the plane for) and might work in your current build.

—Eric

	Eric Pettersen
	UCSF Computer Graphics Lab


> On Mar 1, 2017, at 9:09 AM, Catherine Jenifer Rajam Rajendran <catrajen at umail.iu.edu> wrote:
> 
> Hello,
> 
> When I try to open large protein in Chimera, and try to define plane for the protein, It is throwing me an error saying "Memory Error". Can you please help me with this error?
> 
> Note: Attached a snapshot of the error.
> 
> Thanks,
> Catherine
> <Screenshot.png>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20170301/0fcea974/attachment.html>