[Jmol-users] Changing the color of the residues is very slow

[Carles F. <car...@bs...>]

Hi there,

I'm using an embedded jmol for a project and it behaves quite well,  
especially on Windows/IE, as the JVM seems to be faster.
I have a java applet which loads a PDB file cointaining a DNA sequence  
of about 1000 base pairs. The applet and the PDB load pretty fast  
(about 5 sec) and they are very responsive, I can rotate, zoom, etc.,  
in real time.

Then, I run a script (via a jmolButton, but I think that's irrelevant)  
to change the colour of some residues. The contents of the script is  
something similar to:

select C1, N2, C3, O4, N5, C6, N7, C8, C9, C10, N11, C12, O13, N14,  
C15, O16, C17, C18, C19, C20, N21, C22, O23, N24, C25, O26, C27, C28,  
C29, C30, N31, C32, O33, N34, C35, O36, C37, C38, C39, C40, N41, C42,  
N43, C44, C45, N46, N47, C48, N49, C50, C51, N52, C53, N54, C55, C56,  
N57, N58, C59, N60, C61, C62, N63, C64, N65, C66, C67, N68, N69, C70,  
N71, C72; color [0, 0, 0];
select C73, N74, C75, O76, N77, C78, O79, C80, C81, C82; color [0, 72,  
0];
select C83, N84, C85, O86, N87, C88, O89, C90, C91, C92; color [0,  
115, 0];
...
...

Basically I select every atom, grouped with some others, then change  
their colour. The script file is about 100 Kbytes and it cointains  
about 800 lines, and the internet connection isn't an issue.

My problem is that jmol takes a very very long time (4 minutes for a  
1000 bps DNA, about 10 minutes for 2000 bps) to change the colour of  
all atoms. As the atoms are coloured when the PDB is initially loaded,  
I might be doing it wrong.

Some more info: if I scroll the browser window up and down I can see  
that the applet repaints and, each time, it has more and more atoms  
coloured, but at a very slow rate. During the execution of the script,  
the applet is locked.


Is there any solution to this? I don't really think that changing the  
colour of an atom can take that long. Maybe there is some special  
instruction to do it faster or in a batch?

Thanks a lot!
Carles

-- 
Carles Fenollosa Bielsa
Barcelona Supercomputing Center / Life Sciences
Nexus II / Jordi Girona, 29 / 08034 Barcelona
+34 93 413 76 02