Dear all,
I am new to simulate nucleic acid/ proteins.
I am now trying to simulation t-RNA, the pdb file of it contains some modified residues like 2MG, H2U, M2G, 7MG etc. But their topology files etc are not available in the CHAMMS force fields.
thank you for your reply,
Tim
Dear all,
I am new to simulate nucleic acid/ proteins.
I am now trying to simulation t-RNA, the pdb file of it contains some modified residues like 2MG, H2U, M2G, 7MG etc. But their topology files etc are not available in the CHAMMS force fields.
- So, shall I have to make new topology files by editing the existing ones, if possible? (what can be method for that)?
This is a question better asked in the charmm forums, particularly the one discussing force field parameters.
or
2) is there any data source where I can search for the modified residue topology files etc?
Again, you are asking a charmm force field question.
Furthermore, please note that lammps currently only supports old charmm force field variants up to 22, I think (pre-cross-terms).
I'm a huge fan of LAMMPS, but why are you using LAMMPS for this project?
If you want MD software which can automatically understand what "2MG",
"H2U", "M2G", "7MG" means, well, that's not LAMMPS. (In my opinion,
restricting yourself to windows does not help either.)
http://lammps.sandia.gov/non_features.html
If you are just trying to get a quick project done for a class, there
are many other software programs which would suit your needs much more
quickly. Figuring out how to run your simulation in the current
version of LAMMPS might require more time and effort than it require
using older MD tools. If you are beginning your PhD, perhaps it's
okay to invest this time learning LAMMPS or other new MD codes. (And
some programming would be good too.) But check out the other older
more specialized tools.
I know people frequently use AMBER for nucleic-acid simulations.
there are force-fields appropriate for nucleic acids available, and
there are tutorials and a community of people who can help you (if you
have an AMBER licence).
But, it's not necessary to have an AMBER licence to use the AMBER
force-field. I think mosr amber force-fields are available in GROMACS
& NAMD, which are free, or free for academic use. The same is true
for the CHARMM force-fields. (Another promising choice is OpenMM, but
it's newer, and I suspect it does not yet have as large a community.)
I'm sure there are other choices too.
Perhaps one day, LAMMPS tools will include more features for
all-atom biopolymer simulations. (Who knows, some of us might write
them. And if you want to do that or contribute to that effort, we can
help you!) But I suspect LAMMPS is likely lag behind other MD
distributions which were designed solely for that purpose.
That ended up being a long rant.
(Forgive me for that.)
Cheers
Andrew
Thank you Axel and Andrew for your comments. I am ofcourse taking into account your suggestions.
When I tried to generate the .psf and .pdb file with the existing CHARMM topology files and the downloaded t-RNA pdb file, it worked fine (I think) for all the atoms except for the ‘modified residues’ where it failed. I think it did not work for those modified residues because it could not recognize/import them from the CHARMM topology file (as the modified residues are not enlisted there), that I used.
However, I have found this topology tutorial (http://www.ks.uiuc.edu/Training/Tutorials/science/topology/topology-tutorial.pdf) which demonstrates how in some cases, the topology files for “new residuals” can be formed/generated from the already “existing ones”.
So I shall now try to manually build topology files for the new ‘modified residues’ and will check if it works.
Any comments of yours in this issue is highly appreciated.
thank you,
- Thank you Axel and Andrew for your comments. I am ofcourse taking into account your suggestions.
No, you don’t.
This is not the right forum for this question.
Furthermore, your questions indicate, that you are lacking the experience to immediately tackle the kind of problem you want to simulate. You need to spend some significant time working on simpler systems and also better understand force field paramerization before even considering moving forward on your project, or you are bound to make a lot of mistakes and waste a lot of your and other people’s time.
Of course you are free to do as you like, but don’t expect a lot of support, especially not from a forum that is off-topic.
Axel
...but if you do figure out how to do this, please write back and tell
us how you did it!
(I mean, I'd be curious, and I don't browse the charmm forums.)
Andrew
- Thank you Axel and Andrew for your comments. I am ofcourse taking into
account your suggestions.No, you don't.
This is not the right forum for this question.
Of course you are free to do as you like, but don't expect a lot of support,
especially not from a forum that is off-topic....but if you do figure out how to do this, please write back and tell
us how you did it!
what do you mean by "figure out". the charmm topology file format is
easy enough to understand. the problem is not to add a new entry to
those files, the problem is what numbers to stick in there and how to
confirm that they work correctly. parameterizing force field
components is a complex task that requires equal parts inspiration,
transpiration and black magic. there are a bunch of tools to help with
the mechanics of the process, but that doesn't guarantee an acceptable
result. even people with a lot of experience don't always get it right
the first (or second, third, ...) time.