- Posts: 5
accuracy of dspc results
- Lindén
- Topic Author
- Offline
- Fresh Boarder
Being a martini (and gromacs) newbie about halfway through the tutorial, I have a question about agreement with experimental bilayer data for DSPC. Basically, should I be worried about 10-15% disagreement compared to the data in the Balgavý paper?
Experimental values for DSPC at 333K:
thickness 5.0 nm, area per lipid 0.65 nm^2
My simulated values (after bilayer formation, 15 ns equilibration, and 15 ns production run) at 333K
thickness 4.3 nm (15% too low), area per lipid 0.71 nm^2 (10% too high).
Is this a reasonable agreement, or a sign that I did something wrong?
Sincerely,
Martin Lindén
Please Log in or Create an account to join the conversation.
- xavier
- Offline
- Admin
- Posts: 416
Lindén wrote: Hi!
Being a martini (and gromacs) newbie about halfway through the tutorial, I have a question about agreement with experimental bilayer data for DSPC. Basically, should I be worried about 10-15% disagreement compared to the data in the Balgavý paper?
Experimental values for DSPC at 333K:
thickness 5.0 nm, area per lipid 0.65 nm^2
My simulated values (after bilayer formation, 15 ns equilibration, and 15 ns production run) at 333K
thickness 4.3 nm (15% too low), area per lipid 0.71 nm^2 (10% too high).
Is this a reasonable agreement, or a sign that I did something wrong?
Sincerely,
Martin Lindén
Please Log in or Create an account to join the conversation.
- Lindén
- Topic Author
- Offline
- Fresh Boarder
- Posts: 5
Is that more like the expected agreement?
(Area per lipid was computed as the average xy-area/64 pm bootstrap stderr, and the thickness by fitting two Gaussians to the density histogram of P* atoms, i.e., no attempt at electron density).
Best,
Martin
Please Log in or Create an account to join the conversation.
- xavier
- Offline
- Admin
- Posts: 416
Have fun further.
Lindén wrote: Ah, sorry for bothering you with badly converged data. With 300 ns, I get thickness 4.66 +- 6e-5 nm (6% low) and 0.66 pm 6e-5 nm^2/lipid (1.5% high).
Is that more like the expected agreement?
(Area per lipid was computed as the average xy-area/64 pm bootstrap stderr, and the thickness by fitting two Gaussians to the density histogram of P* atoms, i.e., no attempt at electron density).
Best,
Martin
Please Log in or Create an account to join the conversation.