Re: AMBER: Amber9 analyze timecorr manual ambigous

From: Nicolas Lux Fawzi <fawzin.berkeley.edu>
Date: Thu, 12 Jul 2007 15:11:30 -0700

Hi Professor Case
Thanks for your responses with references, I have those and I'll look
for more details within those on this step of this type of calculation.
Regarding the analyze timecorr, now I'm a little more confused because
with my compiled version of amber9, specifying only vec1 (and no vec2)
works fine as listed on pg 26. I thought it was the pg 260 argument
option list that needed to be updated. I run this ptraj script with
only vec1 specified:

trajin mdcrd_NVE
strip :WAT
vector v150-1H3_6HG :1.H3 corr :6.HG order 2
analyze timecorr vec1 v150-1H3_6HG tstep 1.0 tcorr 5000.0 dplr norm out
vv150-1H3_6HG.out

and i get this output:

Auto-correlation functions, normal type
***** Vector length *****
       <r> <rrig> <1/r^3> <1/r^6>
   12.1561 6.2984 0.0007 0.0000

***** Correlation functions *****
      Time <C> <P2> <1/(r^3*r^3)>
     0.000 1.0000 1.0000 1.0000
     1.000 0.9506 0.9849 0.9746
     2.000 0.9249 0.9758 0.9597
     3.000 0.9062 0.9673 0.9495
     4.000 0.8916 0.9593 0.9404
     5.000 0.8747 0.9508 0.9323
     6.000 0.8652 0.9428 0.9305
     7.000 0.8582 0.9347 0.9300
     8.000 0.8535 0.9266 0.9326
     9.000 0.8420 0.9192 0.9309
    10.000 0.8316 0.9115 0.9283
    11.000 0.8184 0.9036 0.9220
......


Actually, I just ran a test with these two scripts:

trajin mdcrd_NVE
strip :WAT
vector v150-1H3_6HG :1.H3 corr :6.HG order 2
analyze timecorr vec1 v150-1H3_6HG tstep 1.0 tcorr 5000.0 dplr norm out
testscript.vec1.out


trajin mdcrd_NVE
strip :WAT
vector v150-1H3_6HG :1.H3 corr :6.HG order 2
analyze timecorr vec1 v150-1H3_6HG vec2 v150-1H3_6HG tstep 1.0 tcorr
5000.0 dplr norm out testscript.vec1vec2.out

where the only difference is specifying vec2 as the same vector, and I
get the same output:

$ diff testscript.vec1.out testscript.vec1vec2.out
1c1
< Auto-correlation functions, normal type

---
 > Cross-correlation functions, normal type
4a5
 >    12.1561     6.2984     0.0007     0.0000
where the only differences are the header and the additional (identical) 
averages for the vector
David A. Case wrote:
> On Thu, Jul 12, 2007, Nicolas Lux Fawzi wrote:
>
>   
>> 1. I was slightly confused by the amber9 manual ptraj section pg 260 
>> where it says referring to vec1 and vec2 vecnames:
>>
>> "if the two vectors are the same, the result is an autocorrelation function"
>>
>> while it seems clear from the example on pg 261 that you only need to 
>> specify vec1 to get an autocorrelation function:
>>
>> analyze timecorr vec1 v0 tstep 1.0 tcorr 100.0 out v0.out
>>
>> maybe the manual should be updated to reflect that, since the string of 
>> arguments for analyze timecorr makes it seem like vec2 needs to be specified
>>     
>
> Damn.  We got the writeup on p. 260 corrected (compared to the Amber 8
> manual), but missed the fact that the examples are still wrong: both vec1 and
> vec2 are required, even if they are the same.  I'll post an erratum; thanks
> for reporting this.
>
>
>   
>> 2. Can you give me more detail on the difference between drct direct 
>> approach and FFT method?
>>     
>
> See:
>
> %A C. Peter
> %A X. Daura
> %A W.F. vanGunsteren
> %T Calculation of NMR-relaxation parameters for flexible molecules from
> molecular dynamics simulations
> %J J. Biomol. NMR
> %V 20
> %P 297-310
> %D 2001
>
>   
>> 3. If it's not too much trouble, can you point me to some literature 
>> references using this functionality, specifically the dplr keyword (I'm 
>> trying to compute some timecorrelation functions for NMR dynamics 
>> comparisons).
>>
>>     
>
> A couple of reviews:
>
> %A A.G. Palmer, III
> %T NMR characterization of the dynamics of biomacromolecules
> %J Chem. Rev.
> %V 104
> %P 3623-3640
> %D 2004
>
> %A D.A. Case
> %T Molecular Dynamics and NMR Spin Relaxation in Proteins
> %J Acct. Chem. Res.
> %V 35
> %P 325-331
> %D 2002
>
> ...dac
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>   
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu

Received on Sun Jul 15 2007 - 06:07:30 PDT