memory allocate problem

157 views
Skip to first unread message

변진영

unread,
May 21, 2020, 7:32:02 AM5/21/20
to g_mmpbsa Mailing List
Dear g_mmpbas users,

When I run g_mmpbsa with protein-ligand complex that is represented by charmm force field, I met the Cannot allocate memory error
So I extracted the trajectory file from 0 ~ 5ns(half timesize; md production: 10ns) by using the 'gmx trjconv' to reduce the trajectory file size.
And I didn't have problem when I issued the 'gmx rms' or 'gmx gyrate', which the other topic mentioned as a solution
But I still have the same problem. 

How can I solve this problem?

Thank you

Nikhil Maroli

unread,
May 21, 2020, 7:34:37 AM5/21/20
to g_mm...@googlegroups.com
go through the thread, its already discussed here several times.

--
You received this message because you are subscribed to the Google Groups "g_mmpbsa Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to g_mmpbsa+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/g_mmpbsa/c3f479fd-212c-436e-9a18-030f36e1c618%40googlegroups.com.


--
Regards,
Nikhil Maroli

변진영

unread,
May 21, 2020, 8:34:22 PM5/21/20
to g_mm...@googlegroups.com
Could you link the thread..? I can't find the satisfied answer.

2020년 5월 21일 (목) 오후 8:34, Nikhil Maroli <scin...@gmail.com>님이 작성:

변진영

unread,
May 21, 2020, 8:36:17 PM5/21/20
to g_mmpbsa Mailing List


2020년 5월 22일 금요일 오전 9시 34분 22초 UTC+9, 변진영 님의 말:
Could you link the thread..? I can't find the satisfied answer.

2020년 5월 21일 (목) 오후 8:34, Nikhil Maroli <scin...@gmail.com>님이 작성:
go through the thread, its already discussed here several times.

On Thu, May 21, 2020 at 5:02 PM 변진영 <byunj...@gmail.com> wrote:
Dear g_mmpbas users,

When I run g_mmpbsa with protein-ligand complex that is represented by charmm force field, I met the Cannot allocate memory error
So I extracted the trajectory file from 0 ~ 5ns(half timesize; md production: 10ns) by using the 'gmx trjconv' to reduce the trajectory file size.
And I didn't have problem when I issued the 'gmx rms' or 'gmx gyrate', which the other topic mentioned as a solution
But I still have the same problem. 

How can I solve this problem?

Thank you

--
You received this message because you are subscribed to the Google Groups "g_mmpbsa Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to g_mmpbsa+unsubscribe@googlegroups.com.


--
Regards,
Nikhil Maroli

--
You received this message because you are subscribed to the Google Groups "g_mmpbsa Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to g_mmpbsa+unsubscribe@googlegroups.com.

g_mmpbsa Mailing List

unread,
May 22, 2020, 12:35:31 PM5/22/20
to g_mmpbsa Mailing List

Hi,

You might need to check gmemceil (http://rashmikumari.github.io/g_mmpbsa/Polar-solvation-input-parameters.html#gmemceil) and gridspace (http://rashmikumari.github.io/g_mmpbsa/Polar-solvation-input-parameters.html#gridspace) parameters.

Finer the grid, larger memory (also time) is required for the calculation and decreasing gmemceil can be used to tell how much memory is available for calculation.

g_mmpbsa Mailing List

unread,
May 22, 2020, 12:40:36 PM5/22/20
to g_mmpbsa Mailing List

Also, please read this: http://redmine.gromacs.org/issues/2752.
You might be trying to read tpr file generated from different version of Gromacs.


On Thursday, May 21, 2020 at 12:32:02 PM UTC+1, 변진영 wrote:
Reply all
Reply to author
Forward
0 new messages