Search found 1905 matches

by Peter Eastman
Fri Sep 20, 2019 9:32 am
Forum: OpenMM
Topic: Lin & Goddard's two-phase model
Replies: 1
Views: 8

Re: Lin & Goddard's two-phase model

I'm not familiar with that model. OpenMM writes trajectories in DCD format, which is a standard format most analysis codes can read. Or you can use tools like MDTraj to convert it into most other standard trajectory formats.
by Peter Eastman
Fri Sep 20, 2019 9:28 am
Forum: OpenMM
Topic: free energy of solvation
Replies: 1
Views: 7

Re: free energy of solvation

Take a look at Yank. It's for doing alchemical free energy calculations.
by Peter Eastman
Tue Sep 17, 2019 2:11 pm
Forum: OpenMM
Topic: Modifying atom types
Replies: 3
Views: 28

Re: Modifying atom types

Have you looked at Yank ? It's designed for doing alchemical calculations, and it can automate a lot of this for you. Assuming you want to implement all the alchemical calculations yourself, the simplest thing would probably be to start from a Topology that has the carbon and hydrogen. Presumably th...
by Peter Eastman
Tue Sep 17, 2019 12:54 pm
Forum: OpenMM
Topic: no template found for amino acid
Replies: 6
Views: 83

Re: no template found for amino acid

Well, it's getting a lot further now. That's actually a very unusual error. It's a low level error trying to allocate memory on the GPU. What kind of GPU do you have?
by Peter Eastman
Tue Sep 17, 2019 12:50 pm
Forum: OpenMM
Topic: Modifying atom types
Replies: 3
Views: 28

Re: Modifying atom types

Could you elaborate on what you want to do? Systems don't know anything about atom types. They just have energy terms that depend on parameters. Atom types are internal to ForceFields. They're a tool used in selecting the parameter values. They don't get stored anywhere, so they all get forgotten as...
by Peter Eastman
Tue Sep 17, 2019 10:04 am
Forum: OpenMM
Topic: (POSSIBLE BUG): GBn and GBn2 exhibit unexpected behaviors unlike HCT, OBC1, OBC2 in openmm and igb=8 in amber 18
Replies: 17
Views: 220

Re: (POSSIBLE BUG): GBn and GBn2 exhibit unexpected behaviors unlike HCT, OBC1, OBC2 in openmm and igb=8 in amber 18

Thanks! I made the change and wrote a test case for it. With OBC or GBn, I get excellent agreement with energies computed using sander. With GBn2 the agreement is now much better than before, but not as good as the other models. So I'm investigating to try to figure out whether there's still a remai...
by Peter Eastman
Mon Sep 16, 2019 10:24 am
Forum: OpenMM
Topic: (POSSIBLE BUG): GBn and GBn2 exhibit unexpected behaviors unlike HCT, OBC1, OBC2 in openmm and igb=8 in amber 18
Replies: 17
Views: 220

Re: (POSSIBLE BUG): GBn and GBn2 exhibit unexpected behaviors unlike HCT, OBC1, OBC2 in openmm and igb=8 in amber 18

After investigating a bit more, it turns out that only GBn2 uses different parameters for nucleic acids. GBn uses the same parameters for everything (but I'm also told it may not have been tested much with nucleic acids). Do you see large disagreements with Amber for both of them, or only for GBn2?