[gmx-users] g_clustsize

Vitaly Chaban vvchaban at gmail.com
Mon Jan 19 18:45:26 CET 2009


Hi Ran,

Then for what purpose should '-mol' be used?

In my case it's more convenient to have molecule-based clustering rather than
atom-based (in histo-clust.xvg, nclust.xvg, etc). Looking only at the
number of atoms in the cluster it is sometimes very difficult to
determine how many molecules of different types are there. Especially
when there are many clusters and several types of molecules (ions).


Thanks,
Vitaly


RF> Hi,
RF> Use -n without -mol. Clustering will be atom based.

RF> Good luck,
RF> Ran.

RF> Vitaly Chaban wrote:
>> Hi GMX Colleagues,
>>
>> I need a little clarification.
>>
>> I try to calculate a distribution of clusters of ions in a system with
>> several cations and anions and a few solvent molecules. So I just need
>> a number of cluster consisting of ions and the size of the each (how many
>> ions the cluster includes).
>>
>> To do this I use
>> g_clustsize(402) -n index.ndx -mol -s topol.tpr
>> where index.ndx contains only the number of the atoms of the ions.
>>
>> but the programme outputs:
>> Back Off! I just backed up temp.xvg to ./#temp.xvg.1#
>> Reading frame       0 time    0.000   Reading file topol.tpr, VERSION 4.0.2 (single precision)
>> Reading file topol.tpr, VERSION 4.0.2 (single precision)
>> Using molecules rather than atoms. Not reading index file index.ndx
>>
>> Why does it not read the index file with the atom number I want to
>> cluster?
>>
>>
>> When I see for example histo-clust.xvg:
>>
>> @    title "Cluster size distribution"
>> @    xaxis  label "Cluster size"
>> @    yaxis  label "()"
>> @TYPE xy
>>     0     0.000
>>     1    19.531
>>     2     4.583
>>     3     1.707
>>     4     0.647
>>     5     0.259
>>     6     0.060
>>     7     0.140
>>     8     0.048
>>     9     0.054
>>    10     0.020
>>    11     0.022
>>    12     0.000
>> ................
>> ................
>> ................
>>   577    31.096
>>   578    33.457
>>   579    36.982
>>   580    28.942
>>   581    31.311
>>   582    34.850
>>   583    38.401
>>   584    30.307
>>
>>
>> It looks like all the atoms in the system were used to find clusters.
>>
>> And maxclust.xvg:
>>
>> @    title "Max cluster size"
>> @    xaxis  label "Time (ps)"
>> @    yaxis  label "#molecules"
>> @TYPE xy
>>   0.000000e+00         577
>>   1.000000e-01         576
>>   2.000000e-01         565
>>   3.000000e-01         550
>>   4.000000e-01         567
>>   5.000000e-01         572
>>   6.000000e-01         566
>>   7.000000e-01         578
>>   8.000000e-01         567
>>   9.000000e-01         565
>>   1.000000e+00         563
>>   1.100000e+00         579
>>   1.200000e+00         566
>>
>>
>>
>> Please anybody let me know how to find clusters only among those
>> particles I point out in the index file. So I need only the clusters
>> consisting of ions but solvent is not of interest in my case.
>>
>>
>> Thank you in advance,
>> Vitaly
>>
>>
>>
>> ===============================
>> Vitaly V. Chaban, Ph.D.(ABD)
>> School of Chemistry
>> V.N. Karazin Kharkiv National University
>> Svoboda sq.,4, Kharkiv 61077, Ukraine
>> email: chaban at univer.kharkov.ua,vvchaban at gmail.com
>> skype: vvchaban, mob.: +38-097-8259698
>>





More information about the gromacs.org_gmx-users mailing list