From: Giacomo Fiorin (
Date: Tue Oct 04 2022 - 13:31:28 CDT

VMD 1.9.3 is old, and the 1.9.4 alpha releases are rather mature now.

Second, the Windows build for 1.9.3 specifically is a 32-bit executable, so
you are limited in file size (whether 2 GB or 4 GB I don't remember).
1.9.4 has 64-bit builds available; you should be able to read in your file,
provided you use a stride >> 1 to avoid filling up your system RAM with the
DCD file's contents.


On Tue, Oct 4, 2022 at 2:22 PM Efthymiou, Christos <> wrote:

> I am using VMD 1.9.3 on a Windows 10 computer with 32 GB of RAM.
> The command I used for catdcd was: catdcd -o MD_protein.dcd -otype dcd -i
> indexfile -dcd MD.dcd.
> To create the index file, I loaded my PSF and PDB file into VMD and then
> ran this script from the TK console:
> set fp [open indexfile w]
> set a [atomselect top "protein"]
> puts $fp [$a get index] nonewline
> $a delete
> close $fp
> Please let me know if there is anything I can do to solve this crashing
> issue.
> ------------------------------
> *From:* Ashar Malik <>
> *Sent:* Tuesday, October 4, 2022 7:47 AM
> *To:* Efthymiou, Christos <>
> *Cc:* <>
> *Subject:* Re: vmd-l: Maximum Stride Possible for Publishable Results
> ⚠ Caution: External sender
> Please share some more details about the hardware you are using for this
> work (RAM?).
> Additionally what command did you use to desolvate the protein?
> Previously I have used catdcd to desolvate entire trajectories and have
> never had issues.
> Perhaps if you share more, someone here might be able to figure
> something out.
> On Tue, Oct 4, 2022 at 3:28 AM Efthymiou, Christos <
>> wrote:
> Hi,
> I have run some 100 ns simulations and I would like to analyze them with
> VMD/QwikMD. However, each MD.dcd file is around 45 GB and causes VMD to
> crash. I have tried using catdcd to remove the water molecules with a
> stride of 10, but even after running for several hours it does not finish
> creating the modified file. Is there a maximum stride value that is
> reasonable for analyzing simulations? I am concerned about setting a stride
> value that is too high and then the results not being reliable or as
> detailed as necessary, but I cannot find any information regarding a
> consensus or a reasonable range of stride values for analysis.
> I am also open to any suggestions for analyzing larger simulation files in
> VMD/QwikMD. Thank you for any input!