[gmx-users] -b option for g_order

[Beate Griepernau]

David van der Spoel wrote:
> Beate Griepernau wrote:
>> The trajectory has a frame at 46000ps.
>> The headers of the xtc-file are:
>> [...
>>    natoms=     21548  step=  22940000  time=     45880  prec=      1000
>>    natoms=     21548  step=  22945000  time=     45890  prec=      1000
>>    natoms=     21548  step=  22950000  time=     45900  prec=      1000
>>    natoms=     21548  step=  22955000  time=     45910  prec=      1000
>>    natoms=     21548  step=  22960000  time=     45920  prec=      1000
>>    natoms=     21548  step=  22965000  time=     45930  prec=      1000
>>    natoms=     21548  step=  22970000  time=     45940  prec=      1000
>>    natoms=     21548  step=  22975000  time=     45950  prec=      1000
>>    natoms=     21548  step=  22980000  time=     45960  prec=      1000
>>    natoms=     21548  step=  22985000  time=     45970  prec=      1000
>>    natoms=     21548  step=  22990000  time=     45980  prec=      1000
>>    natoms=     21548  step=  22995000  time=     45990  prec=      1000
>>    natoms=     21548  step=  23000000  time=     46000  prec=      1000
>>    natoms=     21548  step=  23005000  time=     46010  prec=      1000
>>    natoms=     21548  step=  23010000  time=     46020  prec=      1000
>>    natoms=     21548  step=  23015000  time=     46030  prec=      1000
>> ...]
>>
>> The promblem seems to be resolved if we cut the large trajectory (0 to
>> 50 ns) in smaller pieces, at least g_order works fine for a .xtc file
>> including only frames from 45 to 48ns.
>
> Is this the downsampled trajectory?
The Headers given are written out from the long trajectory (50ns, every
10th ps).
It is not only a problem  of g_order, but e.g. also of trjconv when
trying  to write e.g. the final nanoseconds of a long trajectory with
the option -b to a new .xtc file. In this case we get a different error
message: 'Fatal error:
Specified frame doesn't exist or file not seekable'. But the frame does
exist definitely.

Beate

>>
>> Beate
>>
>>
>>
>> David van der Spoel wrote:
>>
>>
>>> Beate Griepernau wrote:
>>>
>>>
>>>> Dear all,
>>>>
>>>> I did a simulation of a membrane over a time period of 50ns writing
>>>> out
>>>> coordinates every picosecond.
>>>>
>>>> Next I edited the  .xtc-file to get a new .xtc file keeping only the
>>>> data of every 10th step using trjconv and option -skip.
>>>>
>>>> On this new .xtc-file I applied g_order with different beginning and
>>>> ending times (option -b and -e).
>>>>
>>>> For most beginning and ending times this procedure works well, but for
>>>> beginning times around 46000ps the routine 'g_order' gets stuck:  I
>>>> don't get an error message, but just the notice 'Skipping frame    
>>>> 0time 45890.000' .
>>>> Using later start times > 46100 works well again.
>>>> I checked my .xtc file using several other routines,  e.g. gmxcheck
>>>> and
>>>> trjconv to write a snapshot of the trajectory at the time im question,
>>>> and it worked well.
>>>>
>>>> Does anybody know why this problem occurs and how it can be solved?
>>>>
>>>>
>>>
>>> This is probably not specific for g_order. Does the trajectory have a
>>> frame at 46000?
>>> Could it be that the original framenumbers are still in place (i.e.
>>> step etc.)?
>>> Please run gmxdump on the xtc file and print the headers of the frames
>>> at 46000 and the one before and after and send it to the list.
>>>
>>>
>>>
>>>> Thanks for your help.
>>>>
>>>> Regards,
>>>> Beate
>>>> _______________________________________________
>>>> gmx-users mailing list
>>>> gmx-users at gromacs.org
>>>> http://www.gromacs.org/mailman/listinfo/gmx-users
>>>> Please don't post (un)subscribe requests to the list. Use the www
>>>> interface or send it to gmx-users-request at gromacs.org.
>>>>
>>>>
>>>
>>>
>>
>> _______________________________________________
>> gmx-users mailing list
>> gmx-users at gromacs.org
>> http://www.gromacs.org/mailman/listinfo/gmx-users
>> Please don't post (un)subscribe requests to the list. Use the www
>> interface or send it to gmx-users-request at gromacs.org.
>
>