On 10/18/11 6:02 AM, Matías wrote: > HI, I´m trying to run modeller in multiple cores on a Cluster with > SGE. Do you have an example script to do this? No, since it is highly dependent on your SGE setup. And it depends what you mean by "run modeller". There are two classes in Modeller for running on SGE systems: sge_qsub_job and sge_pe_job. The first uses 'qsub' from within the Modeller run to start an SGE job to do the computation (typically you run Modeller on your head node or some other machine which can 'qsub' to your cluster). The second is designed to be used in a Modeller job which is running on your cluster in a PE; it will allow Modeller to use all the nodes in that PE. A basic usage of the first for comparative modeling would be something like (see also examples/automodel/model-parallel.py): from modeller import * from modeller.automodel import * from modeller.parallel import * j = sge_qsub_job(maxslave=10) env = environ() a = automodel(env, alnfile=...) a.starting_model = 1 a.ending_model = 100 a.use_parallel_job(j) a.make() This will automatically 'qsub' an SGE array job containing 10 tasks when you call a.make(), so that the modeling is split between 10 processors. See also http://salilab.org/modeller/9.10/manual/node438.html Ben Webb, Modeller Caretaker -- modeller-care@salilab.org http://www.salilab.org/modeller/ Modeller mail list: http://salilab.org/mailman/listinfo/modeller_usage