Clustering problem

[Anonymous]

Hi guys.

I met a problem when I was doing the clustering. I excuted this command but it was unexpectedly ended killed and had no result. I’ve 200 pdb files in this folder.

I’ll be really grateful if it can be solved.

$ /home/suuo/Rosetta/main/source/bin/cluster.default.linuxgccrelease -database /home/suuo/Rosetta/main/database -in:file:fullatom -cluster:radius 3 -nooutput -out:file:silent ./cluster.out in:file:s *.pdb

………………………………

……………omit………………

………………………………

protocols.cluster: RESCORING: 6jx4_A_4tz4_C_local_dock_0608.pdb

protocols.cluster: Adding struc: -1612.63

core.import_pose.import_pose: File ‘6jx4_A_4tz4_C_local_dock_0618.pdb’ automatically determined to be of type PDB

protocols.cluster: RESCORING: 6jx4_A_4tz4_C_local_dock_0618.pdb

protocols.cluster: Adding struc: -1606.15

core.import_pose.import_pose: File ‘6jx4_A_4tz4_C_local_dock_0639.pdb’ automatically determined to be of type PDB

protocols.cluster: RESCORING: 6jx4_A_4tz4_C_local_dock_0639.pdb

protocols.cluster: Adding struc: -1607.23

core.import_pose.import_pose: File ‘6jx4_A_4tz4_C_local_dock_0641.pdb’ automatically determined to be of type PDB

protocols.cluster: RESCORING: 6jx4_A_4tz4_C_local_dock_0641.pdb

protocols.cluster: Adding struc: -1606.12

core.import_pose.import_pose: File ‘6jx4_A_4tz4_C_local_dock_0645.pdb’ automatically determined to be of type PDB

protocols.cluster: RESCORING: 6jx4_A_4tz4_C_local_dock_0645.pdb

protocols.cluster: Adding struc: -1611.45

core.import_pose.import_pose: File ‘6jx4_A_4tz4_C_local_dock_0647.pdb’ automatically determined to be of type PDB

protocols.cluster: RESCORING: 6jx4_A_4tz4_C_local_dock_0647.pdb

protocols.cluster: Adding struc: -1609.8

already killed

[Anonymous]

Unfortunately, the Rosetta clustering software isn’t something that you can recover from a stoppage part way through. You basically have to do it all in one go.

The terminology “killed” implies that something external to Rosetta halted the program. Generally speaking, if you didn’t manually stop the program, this is usually the cluster running software, or other such monitoring software on the system. The two most common reasons for such a monitoring program to kill Rosetta is that the run exceeded either the runtime of memory allocation limits specified. If you’re running on a cluster, check that you’ve set the runtime and memory usage for the program properly, and have sufficient amounts for your needs. (This may need some iterative testing.) If you’re running on a login node of a cluster, be aware that administrators often put limits on memory and runtime for anything running there, to encourage people to launch big jobs through the cluster running software.

Even if you’re running on a local machine and not a cluster, it could be that you’re exhausting the memory limits. I know that Linux has the OOM (out of memory) killer, which will kill programs which are exhausting the amount of memory availible on the machine. You may want to monitor the memory usage as the program runs, and see if you’re using up all the availiable memory.

[Author]

Posts