[gmx-users] 'Cannot allocate memory problem' while performing g_cluster

[sunita at chem.iitb.ac.in]

Dear Users,

I am having following error while performing g_cluser.
It is working fine for small time range. But problem appears for large
time fragment e.g. 20 ns .xtc file which has only protein coordinates. I
don't want to do it in fragments. How can I get the actual cluster over
the whole trajectory?


==========================================================================
Last frame      40000 time 20000.000
Allocated 11765880 bytes for frames
Read 20001 frames from trajectory 30ns_protein.xtc
Computing 20001x20001 RMS deviation matrix
# RMSD calculations left: 0

The RMSD ranges from 0.0208186 to 0.414795 nm
Average RMSD is 0.0419409
Number of structures for matrix 20001
Energy of the matrix is 1211.9 nm
WARNING: rmsd minimum 0 is below lowest rmsd value 0.0208186

Back Off! I just backed up rmsd-dist_native.xvg to ./#rmsd-dist_native.xvg.1#
Making list of neighbors within cutoff Fatal error: realloc for nnb[i].nb
(8480 bytes, file g_cluster.c, line 482, nnb[i].nb=0x0x66b2cef0): Cannot
allocate memory
===============================================================================
g_cluster generated the following incomplete log file.

==================================cluster.log==========================
Using gromos method for clustering
Using RMSD cutoff 0.1 nm
The RMSD ranges from 0.0208186 to 0.414795 nm
Average RMSD is 0.0419409
Number of structures for matrix 20001
Energy of the matrix is 1211.9 nm
=======================================================================

Thanks for any suggestion.

regards,
Sunita