Skip to content

Home

Jump to bottom
Antonio Ulloa edited this page Apr 16, 2017 · 17 revisions

Welcome to the LSNM (Large-Scale Neural Modeling) command_line_version wiki!

The LSNM simulator (commmand-line version) can be executed directly on a unix/linux terminal or remotely on the Neuroscience Gateway (NSG) Portal. Please find below instructions for both.

Running LSNM directly on a unix/linux terminal

The following syntax should be used:

python lsnm.py -m model.txt -w weightslist.txt -s script.py -l lsnm_tvb_link.txt

  • lsnm.py is the python script that runs the simulation.
  • weightlist.txt contains a list of individual *.w weight files (all should be included in working directory).
  • script.py is the script of simulation describing all events in the simulated experiment.
  • lsnm_tvb_link.txt is an optional parameter that describes how the LSNM model is embedded into a Virtual Brain (TVB) connectome.

The output is a large number of *.out files in text format that contain time-series of simulated neural and synaptic activity. They are all saved as output files.

Running LSNM on the NSG portal:

  1. Group all input files (model.txt, weightslist.txt, script.py, and all of the individual *.w files describing connections among modules in your network) into a single folder and compress that folder into a zip file.
  2. Log in to your NSG account at www.nsgportal.org
  3. Upload the zipped input data file into your working NSG folder's Data subdirectory.
  4. Push the Create New Task button
  5. Push the Select Tool button and select the LSNM on Commet option
  6. Enter a user-generated identifier for your simulation in the Description field
  7. Push the Select Input Data button and select the zipped file containing input data
  8. Push the Parameters button and enter the name of your simulation script (e.g., script.py) in the Enter Main Input Filename field.
  9. Modify the Maximum Hours to Run field to an estimate of the duration of your simulation. Remember that your job will be terminated when this time elapses (time that the job is waiting in the queue does not count though), so it is better to overestimate running time.
  10. Push the Save and Run Task button
Clone this wiki locally