[gmx-users] Schlitter Configurational entropy 'NaN'

[David van der Spoel]

On 2014-05-15 10:51, tarak karmakar wrote:
>   | Is your system far out of equilibrium?
>
> It should not be as I have simulated the system for 100ns. Few properties
> to check to get the feelings of equilibrium are looking good.
>
> |Do the eigenvalues look reasonable?
> Yes. File is attached.
>
> | Or is your system very large?
> My system is a protein dimer.
>
> It creates the following covariance matrix (c-alpha atoms)
>
> Constructing covariance matrix (1386x1386)
That is big.

Try running g_anaeig -debug 1 and check the output file g_anaeig.debug.
Then check the source code (gmx_anaeig.c) to see whether the evaluation 
of the entropy can be made more numerically stable.
>
> eigenvec.trr file is also attached.
>
> Mail with the attached files
> "Needs moderator approval because of the file size."
>
>
> Thanks,
> Tarak
>
>
> On Thu, May 15, 2014 at 2:19 PM, tarak karmakar <tarak20489 at gmail.com>wrote:
>
>> | Is your system far out of equilibrium?
>>
>> It should not be as I have simulated the system for 100ns. Few properties
>> to check to get the feelings of equilibrium are looking good.
>>
>> |Do the eigenvalues look reasonable?
>> Yes. File is attached.
>>
>> | Or is your system very large?
>> My system is a protein dimer.
>>
>> It creates the following covariance matrix (c-alpha atoms)
>>
>> Constructing covariance matrix (1386x1386)
>>
>> eigenvec.trr file is also attached.
>>
>>
>> Thanks,
>> Tarak
>>
>>
>>
>>
>> On Thu, May 15, 2014 at 2:06 PM, David van der Spoel <spoel at xray.bmc.uu.se
>>> wrote:
>>
>>> On 2014-05-15 10:30, tarak karmakar wrote:
>>>
>>>> Dear Sir,
>>>> Thanks for the quick reply.
>>>> Using g_anaeig_d I've got both of them as 'nan'
>>>>
>>>> The Entropy due to the Quasi Harmonic approximation is -nan J/mol K
>>>> The Entropy due to the Schlitter formula is nan J/mol K
>>>>
>>>
>>> Interesting. Is your system far out of equilibrium? Do the eigenvalues
>>> look reasonable? Or is your system very large?
>>>
>>>
>>>> Tarak
>>>>
>>>>
>>>> On Thu, May 15, 2014 at 1:30 PM, David van der Spoel
>>>> <spoel at xray.bmc.uu.se>wrote:
>>>>
>>>>   On 2014-05-15 09:07, tarak karmakar wrote:
>>>>>
>>>>>   Dear All,
>>>>>> I wanted to calculate configurational entropy of a protein by using
>>>>>> g_covar
>>>>>> and g_anaeig as follows,
>>>>>>
>>>>>> g_covar -f ../traj.xtc -s ../npt_prod -o eigenval -v eigenvec.trr -av
>>>>>> average.pdb
>>>>>> g_anaeig -v eigenvec.trr -entropy -temp 300
>>>>>>
>>>>>> I got the following results
>>>>>>
>>>>>> The Entropy due to the Quasi Harmonic approximation is 31440.3 J/mol K
>>>>>> The Entropy due to the Schlitter formula is nan J/mol K
>>>>>>
>>>>>> I went back to check the eigenvector file by dumping it to a new file
>>>>>> 'dump'. There is no 'nan' indeed.
>>>>>>
>>>>>> Could anyone comment on why I'm getting Schlitter entropy as 'nan'?
>>>>>>
>>>>>> Thanks and regards,
>>>>>> Tarak
>>>>>>
>>>>>>    Numerical error due to large numbers. Try compiling gromacs in double
>>>>>>
>>>>> precision.
>>>>>
>>>>> --
>>>>> David van der Spoel, Ph.D., Professor of Biology
>>>>> Dept. of Cell & Molec. Biol., Uppsala University.
>>>>> Box 596, 75124 Uppsala, Sweden. Phone:  +46184714205.
>>>>> spoel at xray.bmc.uu.se    http://folding.bmc.uu.se
>>>>> --
>>>>> Gromacs Users mailing list
>>>>>
>>>>> * Please search the archive at http://www.gromacs.org/
>>>>> Support/Mailing_Lists/GMX-Users_List before posting!
>>>>>
>>>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>>>
>>>>> * For (un)subscribe requests visit
>>>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>>>>> send a mail to gmx-users-request at gromacs.org.
>>>>>
>>>>>
>>>
>>> --
>>> David van der Spoel, Ph.D., Professor of Biology
>>> Dept. of Cell & Molec. Biol., Uppsala University.
>>> Box 596, 75124 Uppsala, Sweden. Phone:  +46184714205.
>>> spoel at xray.bmc.uu.se    http://folding.bmc.uu.se
>>> --
>>> Gromacs Users mailing list
>>>
>>> * Please search the archive at http://www.gromacs.org/
>>> Support/Mailing_Lists/GMX-Users_List before posting!
>>>
>>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>>>
>>> * For (un)subscribe requests visit
>>> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
>>> send a mail to gmx-users-request at gromacs.org.
>>>
>>
>>
>
>


-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se