choderalab / pymbar

Python implementation of the multistate Bennett acceptance ratio (MBAR)
http://pymbar.readthedocs.io
MIT License
235 stars 91 forks source link

Question about MBAR/parallel-tempering-2dpmf example #256

Open wever opened 7 years ago

wever commented 7 years ago

1 - Plotting free energy estimation seems to me a huge deltaG(target temp=302K), ~5000. Is it as a result from method or is it just a statistical estimation?

2 - I couldn't to reproduce the PMF graph (fig 1. in LONG-TIME PROTEIN FOLDING DYNAMICS FROM SHORT-TIME MOLECULAR DYNAMICS SIMULATIONS) in region (phi=60,psi=...). is it due to different methods (WHAMxMBAR) ?

3 - why k in line: indices = timeseries.subsampleCorrelatedData(U_kt[k,:], g = g)? Seems to me the last k value in last last looping (k=39).

4 - Is it take any possibility about PMF, like Rg x RMSD, number of HB x a-helix percentage..right?

5 - Im working with lysozyme in glycerol. The parameters of glycerol give rise positive potential energy, then, the free energy x temperature decrease to negative values with the augment of temperature. If I calculating differences I didnt expect this behaviour. Do you have some explanation? Thanks.

mrshirts commented 7 years ago

1 - Plotting free energy estimation seems to me a huge deltaG(target temp=302K), ~5000. Is it as a result from method or is it just a statistical estimation?

Is this with a data set of yours? Without knowing what the data set looks like, it's impossible to day.

2 - I couldn't to reproduce the PMF graph (fig 1. in LONG-TIME PROTEIN FOLDING DYNAMICS FROM SHORT-TIME MOLECULAR DYNAMICS SIMULATIONS) in region (phi=60,psi=...). is it due to different methods (WHAMxMBAR) ?

Is there a link to this? It's not quite clear what you are talking about.

3 - why k in line: indices = timeseries.subsampleCorrelatedData(U_kt[k,:], g = g)? Seems to me the last k value in last last looping (k=39).

k=39 sounds like a description of your dataset, but the code is general, so I'm not sure what you are asking.

4 - Is it take any possibility about PMF, like Rg x RMSD, number of HB x a-helix percentage..right?

I think you need to phrase this more precisely -- I'm not sure what you are asking.

5 - Im working with lysozyme in glycerol. The parameters of glycerol give rise positive potential energy, then, the free energy x temperature decrease to negative values with the augment of temperature. If I calculating differences I didnt expect this behaviour. Do you have some explanation? Thanks.

Again, you haven't specified what free energy you are talking about, so I'm not really sure what the issues is. And we definitely don't know that.

wever commented 7 years ago

Its all about of MBAR/parallel-tempering-2dpmf example. figures : https://drive.google.com/drive/folders/0B3Vqft6lxmmCc01wODR1SDRMb2c?usp=sharing

1 - Plot off final dimensionless free energies x temperature (figure 1 link).

2 - figure 2 link

3 - MBAR/parallel-tempering-2dpmf.py example

5 - figure 5 link

Thanks.