Re: [AMBER] SA Bomb in sa_arc()3 error

From: <crazylyf.126.com>
Date: Tue, 9 Jul 2013 08:59:05 +0800 (CST)

Dear Ray,

Thanks a lot!

I have solved this problem by moving it onto our cluster instead of my PC. But there is one thing confusing to me, the program consumes nearly the same memory (~1.4G) for trajectories with different frames. Does this mean that the PBSA package itself require so much memory?

Follow your advice, I tried to modify the SPACE variable from 0.25 to 0.5 and 1.0, and SPACE 1.0 finally works on my PC. That's great, Thank you very much.

Best,

Yufeng



At 2013-07-09 01:23:29,"Ray Luo, Ph.D." <ray.luo.uci.edu> wrote:
>Yufeng,
>
>Yes, this is due to the lack of main memory on your computer. Can you
>try to double the grid spacing (space) and try again? Doing so would
>reduce the memory usage by a factor of 8. If it works, you can try a
>number less aggressive.
>
>All the best,
>Ray
>
>On Mon, Jul 8, 2013 at 1:34 AM, crazylyf <crazylyf.126.com> wrote:
>> Hi everybody,
>>
>> I'm trying to estimate the free energy evolution of a membrane protein from trajectory, using PBSA implicit membrane and solvent in sander (version AmberTools12). After solved several problems, I finally got stuck with such an error:
>> "SA Bomb in sa_arc()3: Allocates aborted 5014 5014 5014 5014 5014"
>>
>> I noticed that previous discussions in the archive pointed out that it may be caused by the lack of memory, and suggested to increase the arcres. However, my calculation still failed with arcres 0.125, 0.25, or even larger. I also tried to split the trajectory into small small pieces, but they aborted quickly again.
>>
>> Could anyone tell me what is going on? Thanks in advance!
>>
>> Best,
>>
>> Yufeng
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>
>_______________________________________________
>AMBER mailing list
>AMBER.ambermd.org
>http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jul 08 2013 - 18:30:02 PDT
Custom Search