Re: Re: vmd-l: Fwd: FEP with intermediary segments

From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Sat May 05 2018 - 01:21:23 CDT

Options for ParseFEP:
-- Gram-Charlier order 0
v disp
v entropy
v Gaussian approximation
FEP output file
FEP(bachward) output file
Combine frwd ancd back sampling
BAR
______________
ParseFEP for "frwd 0.20 0.60 0.1" went on regularly, with graphic outputs.

With "back 0.60 0.20 -0.1", both frwd and back FEP ended OK (for example
> the last part of the log for frwd:
> WRITING EXTENDED SYSTEM TO RESTART FILE AT STEP 6000000
> WRITING EXTENDED SYSTEM TO OUTPUT FILE AT STEP 6000000
> CLOSING EXTENDED SYSTEM TRAJECTORY FILE
> WRITING COORDINATES TO OUTPUT FILE AT STEP 6000000
> CLOSING COORDINATE DCD FILE frwd-03.dcd
> WRITING VELOCITIES TO OUTPUT FILE AT STEP 6000000
> ====================================================
>
> WallClock: 59221.542969 CPUTime: 59221.542969 Memory: 591.234375 MB
> [Partition 0][Node 0] End of program
>

Also fepout looked like OK (for example for frwd:

# STEP Elec
> vdW dE dE_avg Temp dG
> # l l+dl l
> l+dl E(l+dl)-E(l)
> #NEW FEP WINDOW: LAMBDA SET TO 0.6 LAMBDA2 0.7
> FepEnergy: 2000 -9244.0498 -9244.0504 821.6972
> 824.6200 2.9223 2.3404 293.6815 1.1972
> FepEnergy: 4000 -9328.6470 -9328.6476 822.9571
> 825.2196 2.2619 2.4013 303.1666 1.4052
> FepEnergy: 6000 -9428.4316 -9428.4321 930.1513
> 930.8992 0.7474 2.2119 298.6275 1.2794
> .............................
> ................................
>
> FepEnergy: 5998000 -9505.1134 -9505.1129 934.5882
> 935.3368 0.7492 1.1524 307.3336 -5.2350
> FepEnergy: 6000000 -9322.3841 -9322.3836 847.8147
> 851.0300 3.2158 1.1516 291.3593 -5.2347
> #Free energy change for lambda window [ 0.9 1 ] is -5.23473 ; net change
> until now is -3.13571
>

However, ParseFEP gave no graphics for combined frwd-back.
With frwd alone, also graphic output was obtained: attached png for summary

Hope to have provided relevant information enough

francesco

On Fri, May 4, 2018 at 10:46 PM, Brian Radak <brian.radak_at_gmail.com> wrote:

> I mean what were the options you gave to ParseFEP? It would also be useful
> to see portions of the output which summarize the data.
>
>
> On Fri, May 4, 2018, 4:00 PM Francesco Pietra <chiendarret_at_gmail.com>
> wrote:
>
>> The VMD forum was omitted to avoid irritating some folks (I am not
>> referring to Alice Cooper's nightmare).
>>
>> The calculation for the Unbound was carried out as I described:
>>
>> frwd 0.00 0.20 0.05
>>> back 0.20 0.00 -0.05
>>>
>>> using preeq 1ns, FEP 6ns, and I/O outputs 2000
>>>
>>> Because of scalability problems, I have now divided the remaining region
>>> in two sectors.
>>> For the 0.2-0.6 sector:
>>>
>>> frwd 0.20 0.60 0.1
>>>
>> back 0.60 0.20 -0.1
>>
>>
>> using the layout of the .namd input files as in the tutorial. Should that
>> be not enough, please instruct me how to satisfy your request.
>>
>> thanks
>>
>> francesco
>>
>>
>>
>> On Fri, May 4, 2018 at 6:04 PM, Brian Radak <brian.radak_at_gmail.com>
>> wrote:
>>
>>> That's step one - how did you run your calculation?
>>>
>>> On Fri, May 4, 2018 at 11:51 AM, Francesco Pietra <chiendarret_at_gmail.com
>>> > wrote:
>>>
>>>> VMD 1.9.4a12
>>>>
>>>> ParseFEP was provided to me fro Dr Chris Chipot on last Mar 8.
>>>> From that:
>>>> # Chris Chipot, Liu Peng, 2007-2012
>>>> # INCLUSION OF PARSER. MAY 28, 2009, C.C.
>>>> # 1.8 2013-12-20
>>>> #----------------------------------------------------------------------
>>>>
>>>> # $Id: parseFEP.tcl,v 1.14 2014/12/10 16:08:57 johns Exp $
>>>> #----------------------------------------------------------------------
>>>>
>>>> # YingChih Chiang (Letitia), 2017/01/18, 2017/11/20
>>>> # Re-structure the source code for performance:
>>>> # (1) Reading fepout files are done via calling readfepout.
>>>> # (2) Only read the fepout files once and save the data.
>>>> # (3) Bugs of reading fepout files in BAR_estimator and
>>>> # inaccuracy_estimation have been removed.
>>>> # YingChih Chiang (Letitia), 2018/02/01, 2018/02/10
>>>>
>>>> francesco
>>>>
>>>> On Fri, May 4, 2018 at 3:01 PM, Brian Radak <brian.radak_at_gmail.com>
>>>> wrote:
>>>>
>>>>> This sounds like a question regarding ParseFEP. Can you be more
>>>>> specific about:
>>>>>
>>>>> - which version of ParseFEP/VMD you are using?
>>>>> - what options you used (I presume from the GUI, not the terminal)?
>>>>>
>>>>> On Fri, May 4, 2018 at 1:37 AM, Francesco Pietra <
>>>>> chiendarret_at_gmail.com> wrote:
>>>>>
>>>>>> Perhaps this question about "FEP with Intermediary Segmants" has
>>>>>> passed unnoticed.
>>>>>> thanks
>>>>>> for paying attention to
>>>>>>
>>>>>> fp
>>>>>> ---------- Forwarded message ----------
>>>>>> From: Francesco Pietra <chiendarret_at_gmail.com>
>>>>>> Date: Thu, Apr 26, 2018 at 5:20 PM
>>>>>> Subject: FEP with intermediary segments
>>>>>> To: NAMD <namd-l_at_ks.uiuc.edu>, VMD Mailing List <vmd-l_at_ks.uiuc.edu>
>>>>>>
>>>>>>
>>>>>> Hello:
>>>>>> For the Unbound ligand, in the frame of ligand-receptor FEP, I
>>>>>> carried out, formally successfully, the following segment of the
>>>>>> Hamiltonian steep region
>>>>>>
>>>>>> frwd 0.00 0.20 0.05
>>>>>> back 0.20 0.00 -0.05
>>>>>>
>>>>>> using preeq 1ns, FEP 6ns, and I/O outputs 2000
>>>>>>
>>>>>> Because of scalability problems, I have now divided the remaining
>>>>>> region in two sectors.
>>>>>> For the 0.2-0.6 sector:
>>>>>>
>>>>>> frwd 0.20 0.60 0.1
>>>>>> back 0.60 0.20 -0.1
>>>>>>
>>>>>> ParseFEP gave only a log file with tables for
>>>>>> Free energy perturbation
>>>>>> Inaccuracy
>>>>>> perturbation theory
>>>>>>
>>>>>> NO graphic outputs and no useful warning
>>>>>>
>>>>>> Was the setting up of the FEP incorrect, or intermediary sectors can
>>>>>> not be used?
>>>>>>
>>>>>> Thanks for advice
>>>>>> francesco pietra
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>


summary.png

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:21:04 CST