CHARMM Development Project Forums User Discussion & Questions Parameter Set Discussion Nucleic Acid Stains

Dear all,

i want to simulate the passing of some nucleic acids stains through a certain ion channel, and i was wondering if anyone had the parameters for any of them.
The ones I need are these:

-Acridine orange
-DAPI
-ethidium bromide
-ethidium homodimer-2
-evans blue
-hexidium iodide
-indocyanine green
-methylene blue
-propidium iodide
-nuclear yellow

I'm asking only in case someone already has any of these, since i'm going to try generating these parameters with the CGenFF webserver.


Thanks!

You have to be careful; these appear to be dyes, and a lot of dyes have a net charge that is strongly delocalized over a long distance. This can be determined by comparing the pKa to the pH at which you wish to simulate, and if the charged species is predicted to be dominant, writing down resonance structures. If a large-scale charge delocalization is indicated, the Charmm General Force Field is not optimally equipped to deal with it, and even automatic parametrization tools like ffTK may have difficulties. I'm not going to look up the structures and pKas of all the molecules you listed, but if you are in doubt about one or more of these cases, feel free to ask more specific questions.

That's correct, Kenno: all of them are fluorescent dyes.
As you suggested through the NAMD mailing list (thanks for answering!), i already submitted them to the CGenFF program and got not-so-good results, since for almost every molecule there're plenty of penalty scores higher than 100.0 (especially for charges and dihedrals).

Still, I'll take a look at the pKa of each molecule and make sure i submitted the dominant species (i'm simulating at neutral pH).

Thanks!

yes, submitting the dominant species is very important, but please do check on the large-scale charge delocalization too, because the program is likely to give qualitatively incorrect results if your molecule has that specific feature. This is a known limitation that has its roots in the coverage of the underlying force field. That's not to say we cannot address it; it 's just that it's a big effort and we have other priorities right now. Similarly, as far as you are concerned, it is feasible to parametrize such molecules, but they will take you far more time and effort than the ones that don't have this feature, so I would suggest keeping them for last and recruiting help when you eventually get to them.