[Dock-fans] out of memory mallocing in PBSA

Scott Brozell sbrozell at scripps.edu
Mon Jun 2 13:22:45 PDT 2008


On Mon, 2 Jun 2008, Prashant Kumar wrote:

>                                I am using DOCK6.1 with zap for MM-PBSA
> calculation . While performing the docking in rigid manner I am getting good
> results with no problem , but in other hand  for flexible one I am getting
> following  error .
> Initializing Library File Routines...
> Initializing Conformer Generator Routines...
> Initializing PB/SA Score Routines...
> Initializing Grid Score Routines...
>  Reading the energy grid from /home/prashant/olddock/1KZK/3_grid/grid1.nrg
> out of memory mallocing  31786900 for Eps.x!

The last message is from the ZAP library.

> *******
> Total oe_memory currently allocated=        0.552148 Mbytes
> Total oe_object_memory currently allocated= -1245.293528 Mbytes
> Total memory currently allocated=           -1244.741380 Mbytes
> *******pointer/memory list*******
> pointer -1235980280 has 262144 bytes of memory

The negative values suggest that something has overflowed.

> I want to clear one thing in my case rigid docking is not possible . I also
> did some things with ulimit unix command but nothing worked out . Please
> give me some suggestion to rectify this problem .

The general recipe for these resource exhaustion situations is to
benchmark your system: start with a small calculation (small grid,
small receptor, ...) and systematically increase its size.
Eventually, you will develop a memory consumption model as a function
of the various independent variables above (grid size, ...).

Get rich quick schemes involve doing some arithmetic
(how much total memory on your platform, how much available memory,
how big are the grid files) in the hopes finding a big memory sink
or trying a different platform in the hopes of finding more resources.

Look at threads involving Francesco Pietra over the last few months
for other tidbits.

Good luck,

More information about the Dock-fans mailing list