Hello, Is there any plan to plug some alternative molecular surface calculation algorithm into chimera? I have seen some interesting articles about isosurface and gaussian density functions: "Yu, Z., Jacobson, M. P. and Friesner, R. A. (2006), What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules. J. Comput. Chem., 27: 72–89. doi: 10.1002/jcc.20307" "What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules DOI: 10.1002/jcc.20307" http://onlinelibrary.wiley.com/doi/10.1002/jcc.20307/abstract The algorithm looks simple, robust and even parallelizable. I kind of guess that's the algorithm used in MOE when I look at their triangulation. Regards, Francois.
We are starting work on Chimera 2, which will replace the aging user interface (Tk) and legacy graphics library calls (OpenGL 1 and 2). We will probably replace MSMS as part of that work. We have examined a number of alternative algorithms, most of which do not compute analytical solutions as MSMS does. However, the numerical algorithms seem acceptable and are more efficient, so we will probably go with one of them. If there is a favorite molecular surface algorithm that you'd like to see in Chimera, please let us know (both which algorithm and why, which would greatly help us evaluate their relative strengths and weaknesses). Conrad On 2/19/13 6:10 PM, Francois Berenger wrote:
Hello,
Is there any plan to plug some alternative molecular surface calculation algorithm into chimera?
I have seen some interesting articles about isosurface and gaussian density functions:
"Yu, Z., Jacobson, M. P. and Friesner, R. A. (2006), What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules. J. Comput. Chem., 27: 72–89. doi: 10.1002/jcc.20307"
"What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules DOI: 10.1002/jcc.20307" http://onlinelibrary.wiley.com/doi/10.1002/jcc.20307/abstract
The algorithm looks simple, robust and even parallelizable. I kind of guess that's the algorithm used in MOE when I look at their triangulation.
Regards, Francois. _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
Hi Conrad, I'm sure that you are much better versed in and better positioned to chose than I. However, here are some thoughts: First, is it critical that the mesh be triangular? Or can it be made of quads or n-gons? How important is the regularity of the mesh? A Voronoi surface implementation would be interesting, but it might present more difficulties than it is worth when trying to map bio/chem/physical properties to the surface. VMD uses isosurfaces extracted from volumetric Gaussian density maps ("quicksurf") to create surfaces fast enough to be recalculated on the fly for animating trajectories. This is similar to the tip that Elaine and Tom posted about "faking" a molecular surface when MSMS fails: http://www.ks.uiuc.edu/Research/vmd/current/ug/node72.html http://plato.cgl.ucsf.edu/pipermail/chimera-users/2012-January/007048.html The BioBlender project calls PyMOL for generation of its surfaces. As you know, PyMOL surfaces are distinctive and attractive, but not much faster than MSMS: http://bioblender.eu http://www.scivis.ifc.cnr.it/images/stories/PDFarticle/1009.4674.pdf http://www.mail-archive.com/pymol-users@lists.sourceforge.net/msg06658.html I like the spherical harmonics approach of Bruce Duncan, but I suspect that Chimera already incorporates this technique in the "Multiscale Models" module. http://mgl.scripps.edu/people/bsd/cv/cv.html I wonder if "Surface Racer" could be implemented somehow. It seems very fast and it calculates per-residue surface areas, just like the current MSMS implementation: http://apps.phar.umich.edu/tsodikovlab/index_files/Page756.htm OK, I'm tapped out for now... Cheers, Darrell Darrell Hurt, Ph.D. Section Head, Computational Biology Bioinformatics and Computational Biosciences Branch (BCBB) OCICB/OSMO/OD/NIAID/NIH 31 Center Drive, Room 3B62B, MSC 2135 Bethesda, MD 20892-2135 Office 301-402-0095 Mobile 301-758-3559 http://bioinformatics.niaid.nih.gov (Within NIH) http://exon.niaid.nih.gov (Public) Disclaimer: The information in this e-mail and any of its attachments is confidential and may contain sensitive information. It should not be used by anyone who is not the original intended recipient. If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage devices. National Institute of Allergy and Infectious Diseases shall not accept liability for any statements made that are sender's own and not expressly made on behalf of the NIAID by one of its representatives. On 2/20/13 12:24 PM, "Conrad Huang" <conrad@cgl.ucsf.edu> wrote:
We are starting work on Chimera 2, which will replace the aging user interface (Tk) and legacy graphics library calls (OpenGL 1 and 2). We will probably replace MSMS as part of that work.
We have examined a number of alternative algorithms, most of which do not compute analytical solutions as MSMS does. However, the numerical algorithms seem acceptable and are more efficient, so we will probably go with one of them. If there is a favorite molecular surface algorithm that you'd like to see in Chimera, please let us know (both which algorithm and why, which would greatly help us evaluate their relative strengths and weaknesses).
Conrad
On 2/19/13 6:10 PM, Francois Berenger wrote:
Hello,
Is there any plan to plug some alternative molecular surface calculation algorithm into chimera?
I have seen some interesting articles about isosurface and gaussian density functions:
"Yu, Z., Jacobson, M. P. and Friesner, R. A. (2006), What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules. J. Comput. Chem., 27: 7289. doi: 10.1002/jcc.20307"
"What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules DOI: 10.1002/jcc.20307" http://onlinelibrary.wiley.com/doi/10.1002/jcc.20307/abstract
The algorithm looks simple, robust and even parallelizable. I kind of guess that's the algorithm used in MOE when I look at their triangulation.
Regards, Francois. _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
On 02/21/2013 05:44 AM, Hurt, Darrell (NIH/NIAID) [E] wrote:
Hi Conrad,
I'm sure that you are much better versed in and better positioned to chose than I. However, here are some thoughts:
First, is it critical that the mesh be triangular? Or can it be made of quads or n-gons? How important is the regularity of the mesh? A Voronoi surface implementation would be interesting, but it might present more difficulties than it is worth when trying to map bio/chem/physical properties to the surface.
VMD uses isosurfaces extracted from volumetric Gaussian density maps ("quicksurf") to create surfaces fast enough to be recalculated on the fly for animating trajectories.
I think that's what MOE does also. That's a wise choice.
This is similar to the tip that Elaine and Tom posted about "faking" a molecular surface when MSMS fails: http://www.ks.uiuc.edu/Research/vmd/current/ug/node72.html http://plato.cgl.ucsf.edu/pipermail/chimera-users/2012-January/007048.html
The BioBlender project calls PyMOL for generation of its surfaces. As you know, PyMOL surfaces are distinctive and attractive, but not much faster than MSMS: http://bioblender.eu http://www.scivis.ifc.cnr.it/images/stories/PDFarticle/1009.4674.pdf http://www.mail-archive.com/pymol-users@lists.sourceforge.net/msg06658.html
If I can vote, then I vote against Pymol's surface. It's just a hack, brute-force and inefficient algorithm (according to Waren Delano himself when you look into the subject on the net).
I like the spherical harmonics approach of Bruce Duncan, but I suspect that Chimera already incorporates this technique in the "Multiscale Models" module. http://mgl.scripps.edu/people/bsd/cv/cv.html
I wonder if "Surface Racer" could be implemented somehow. It seems very fast and it calculates per-residue surface areas, just like the current MSMS implementation: http://apps.phar.umich.edu/tsodikovlab/index_files/Page756.htm
OK, I'm tapped out for now...
Cheers, Darrell
Darrell Hurt, Ph.D. Section Head, Computational Biology Bioinformatics and Computational Biosciences Branch (BCBB) OCICB/OSMO/OD/NIAID/NIH
31 Center Drive, Room 3B62B, MSC 2135 Bethesda, MD 20892-2135 Office 301-402-0095 Mobile 301-758-3559 http://bioinformatics.niaid.nih.gov (Within NIH) http://exon.niaid.nih.gov (Public)
Disclaimer: The information in this e-mail and any of its attachments is confidential and may contain sensitive information. It should not be used by anyone who is not the original intended recipient. If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage devices. National Institute of Allergy and Infectious Diseases shall not accept liability for any statements made that are sender's own and not expressly made on behalf of the NIAID by one of its representatives.
On 2/20/13 12:24 PM, "Conrad Huang" <conrad@cgl.ucsf.edu> wrote:
We are starting work on Chimera 2, which will replace the aging user interface (Tk) and legacy graphics library calls (OpenGL 1 and 2). We will probably replace MSMS as part of that work.
We have examined a number of alternative algorithms, most of which do not compute analytical solutions as MSMS does. However, the numerical algorithms seem acceptable and are more efficient, so we will probably go with one of them. If there is a favorite molecular surface algorithm that you'd like to see in Chimera, please let us know (both which algorithm and why, which would greatly help us evaluate their relative strengths and weaknesses).
Conrad
On 2/19/13 6:10 PM, Francois Berenger wrote:
Hello,
Is there any plan to plug some alternative molecular surface calculation algorithm into chimera?
I have seen some interesting articles about isosurface and gaussian density functions:
"Yu, Z., Jacobson, M. P. and Friesner, R. A. (2006), What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules. J. Comput. Chem., 27: 7289. doi: 10.1002/jcc.20307"
"What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules DOI: 10.1002/jcc.20307" http://onlinelibrary.wiley.com/doi/10.1002/jcc.20307/abstract
The algorithm looks simple, robust and even parallelizable. I kind of guess that's the algorithm used in MOE when I look at their triangulation.
Regards, Francois. _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
http://www.mail-archive.com/pymol-users@lists.sourceforge.net/msg06658.ht ml
If I can vote, then I vote against Pymol's surface. It's just a hack, brute-force and inefficient algorithm (according to Waren Delano himself when you look into the subject on the net).
Francois is certainly correct. The link I cited in my first email is from Warren and in it, he freely admitted that it is a "brute-force" approach. I pointed out PyMOL simply because surfaces from PyMOL look different than surfaces in any other viewer I know of. Despite their drawbacks (computational and scientific), PyMOL surfaces tend to be visually compelling. And given than PyMOL has been used so much in the literature, maybe there is something to that.
On 02/21/2013 02:24 AM, Conrad Huang wrote:
We are starting work on Chimera 2, which will replace the aging user interface (Tk) and legacy graphics library calls (OpenGL 1 and 2). We will probably replace MSMS as part of that work.
At least, there should be alternatives to MSMS.
We have examined a number of alternative algorithms, most of which do not compute analytical solutions as MSMS does. However, the numerical algorithms seem acceptable and are more efficient, so we will probably go with one of them. If there is a favorite molecular surface algorithm that you'd like to see in Chimera, please let us know (both which algorithm and why, which would greatly help us evaluate their relative strengths and weaknesses).
The first one I referenced looks ideal (gaussian, isosurface and marching cubes). I think I saw some papers where people derived analytical formulas to compute surface and volume fast for molecules describbed with gaussians. If I was to do the job, I would run far away from Connolly-style surfaces (because it is hell to implement and you need CGAL to implement it if you want something robust).
Conrad
On 2/19/13 6:10 PM, Francois Berenger wrote:
Hello,
Is there any plan to plug some alternative molecular surface calculation algorithm into chimera?
I have seen some interesting articles about isosurface and gaussian density functions:
"Yu, Z., Jacobson, M. P. and Friesner, R. A. (2006), What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules. J. Comput. Chem., 27: 72–89. doi: 10.1002/jcc.20307"
"What role do surfaces play in GB models? A new-generation of surface-generalized born model based on a novel gaussian surface for biomolecules DOI: 10.1002/jcc.20307" http://onlinelibrary.wiley.com/doi/10.1002/jcc.20307/abstract
The algorithm looks simple, robust and even parallelizable. I kind of guess that's the algorithm used in MOE when I look at their triangulation.
Regards, Francois. _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
_______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users
participants (3)
-
Conrad Huang -
Francois Berenger -
Hurt, Darrell (NIH/NIAID) [E]