Re: AMBER: PTRAJ:"hbond" analysis

From: Vineet Pande <>
Date: Thu, 31 Mar 2005 09:17:57 +0000

Dear Prof. Cheatham:
Thanks a lot for your detailed reply. I think now its much better to
understand for me whats going on.

>From: "Thomas E. Cheatham, III" <>
>Subject: Re: AMBER: PTRAJ:"hbond" analysis
>Date: Wed, 30 Mar 2005 10:40:16 -0700 (Mountain Standard Time)
> > I Used Ptraj for H-bond analysis. I have the following doubts >>>
>Doubts or results?
> > donor mask :63.OE2
> > acceptor WAT O H1
> > acceptor WAT O H2
> >
> > hbond series HBOND6B time 2 \
> > out wat_hbond_63OE2.dat solventneighbor 5 \
> > solventacceptor WAT O H1 solventacceptor WAT O H2 \
>OK, so you are looking at the interaction of *specific* waters (via the
>acceptor command) and general (indistinguishable) waters with the
>solventacceptor command.
> > (1) The distances look OK, but ALL the angles are too low. Since I
> > used a cutoff of 120 deg (default), why these 20, 30, 40 etc.
> > degrees are showing up?
>By default, the distance cutoff is 3.5 A and the angle is 120.0 degrees,
>but this is an admittedly funny definition of the angle as I currently
>look back at my code.
>The angle is calculated as the angle between the
> donor -- acceptorH -- acceptor
>If you had a linear h-bond, this angle would be 180.0. The angle cutoff
>is, at 120.0, stating that the angle formed must be greater than 120.0.
>However, the angle *stored*, for some reason is 180.0 - angle, such that a
>linear hbond would be zero degrees. Therefore, a value of ~30-40 is a
>good hbond...
> > (2) What does that last column mean :
> > |.-xxo**o |
>This is a schematic of occupancy as a function of time from the beginning
>to the end of the trajectory; the darker the symbol (loosely), the greater
>the occupancy; at the top of the .dat file you will see the definitions,
>but basically for this case...
> |.-xxo**o |
> show partial occupancy at the beginning, getting larger and then
>going away at the end of the trajectory (i.e. it is no longer formed).
> > (3) What's the difference between "%occupied", "lifetime" and
> > "maxocc"....?
> >
> > (4) Reading the 1st and 5th rows:
> >
> | 45.83 2.699 (0.11) 17.10 ( 8.78) 17.2 ( 13.0) 68 |.-xxo**o
>%occupied is the percent of time the hbond is formed over the trajectory
>(i.e. 45.83%). 2.699 and 17.10 are the distances and angles (along with
>stddev in parenthesis). Lifetime is a measure of the average time the
>hbond is formed, i.e. we sum up all the times from intervals where
>hbond is present and divide by the number of intervals, so this basically
>states that you average lifetime (or amount of time the h-bond was formed)
>at 17.2 ps (assuming you set the time variable correctly). Max occupied
>is the maximum number of (contiguous) frames (minus instantaneous gaps)
>the h-bond was occupied; in your case this is 68 frames * 2 ps or 136 ps.
> > IS IT SAFE TO SAY that WAT9563 showed 68+19=87% maximum occupancy
> > and 45.83+18.83=64.66% occupancy.....if its correct, then what's the
> > difference between two options.
>Yes, according to your definition of a h-bond (i.e. 3.5 A cutoff and 120
>degree angle criteria) the single water (interacting via H1 or H2) is
>occupied 87% of the time.
> > (5) Reading the ultimate last row:
> > solvent donor | 5746 :363.H1 5745 :363.O | 156.83 2.695 (
> > 0.12) 16.56 ( 9.18) 10.6 ( 11.0) 68 |.@@@@@@@@@|
>This is more tricky and will depend your definition of solventdonor (which
>was not in your script/command). Solventdonor provides a way to look at
>the interaction of any of a list of waters or ions or what not with your
>If you assume that solventdonor above was WAT O, this states that a water
>interacted with residue :363 156.8% of the time. How can this be? More
>than a single water can satisfy the criteria; the maximum number that can
>interact with a given donor/acceptor is given by the solventneighbor
>value. If you blow this limit, the program will warn you.
>If you want more information, ad naseum, set prnlev 5 before running
>hbond. It is a complicated bit of code but intends to be general.
>You could define solventdonor Na+ Na+ solventacceptor Na+ Na+ Na+ to look
>at ion interaction with a residue (ignoring the angle cutoff since it
>doesn't make sense), etc.
>Good luck.
>The AMBER Mail Reflector
>To post, send mail to
>To unsubscribe, send "unsubscribe amber" to

Don't just search. Find. Check out the new MSN Search!

The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Thu Mar 31 2005 - 10:53:00 PST
Custom Search