From: Maxim Belkin (mbelkin_at_ks.uiuc.edu)
Date: Thu Jun 26 2014 - 09:43:23 CDT

Hi Puspita,

Getting rid of water greatly speeds up any analysis if/when you don't care about water.

It's easy to prepare what you need with VMD and catdcd.

set sel [atomselect top "not water"]
set out [open "indices_nowater.txt" w]
puts $out "[$sel list]"
close $out
$sel writepsf nowater.psf

catdcd -o nowater.dcd -otype dcd -i indices_nowater.txt trajectory.dcd

At this point, you can load your system without water (mol load psf nowater.psf dcd nowater.dcd) and proceed with the proper analysis.

My 30 Gb trajectories shrink to 2 Gb ones and it's much easier to deal with them.

Maxim

On Jun 26, 2014, at 9:15 AM, Puspita Halder <puspitah_at_gmail.com> wrote:

> Hi,
>
> Thanks for your comments. I did not strip out the water molecules of
> my dcd file. Do u think that would help analyzing the secondary
> structure of the big dcd file using sscache? I am carrying out other
> analyses (e.g., rmsd, radius of gyration, sasa etc.) of my large dcd
> file using bigdcd.tcl. So I was wondering if sscache can be utililized
> for secondary structure determination of the large dcd file aong with
> the bigdcd script. Any view on this would be really helpful.
>
> Thanks
> Puspita
>
> On 6/26/14, Jeremiah Babcock <zhc605_at_my.utsa.edu> wrote:
>> Hi Puspita,
>> Not sure if you did this but if you just need the secondary structure
>> of the protein you could strip out the solvent molecules. That would really
>> help on the file size. Also, which OS version are you using?
>>
>> Regards,
>> Jeremiah Babcock
>>
>>
>> On Wed, Jun 25, 2014 at 8:23 AM, Puspita Halder <puspitah_at_gmail.com> wrote:
>>
>>> Dear VMD users,
>>>
>>> I need to generate the secondary structure of my protein from a 50 ns
>>> simulation run. I used sscache earlier for 2-4 ns simulations of the
>>> protein. Now the file sizes are huge (~ 50 GB) so I assume that I cannot
>>> directly use sscache. I tried to combine sscache with bigdcd.tcl script
>>> but
>>> failed to do so. Please help me with your valuable suggestions regarding
>>> this. Is there any ready script available ?
>>>
>>> Thanks
>>> Puspita
>>>
>>