Experimental protocols

From NeuralNetoff
Jump to: navigation, search

PRC

Setting up Dyanmic clamp for PRC

  1. log into Linux machine
  2. change to root: su - root
  3. cd to rtxi
  4. Create entry into database with Dynamic clamp, choose cell type and get cell#, Module=PRC
  5. load settings file: Ch1PRC_EPID.set
  6. name file /usr/src/rtxi/Data/Date_PRC_Cell#_run#.prc
  7. Before performing PRC characterize cell using Istep naming file Date_Step_Cell#_run#.step, record this under notes in database entry for PRC file
  8. Turn on spike rate controller and let spike rate stabilize. If cell is spiking doublets adjust spike detector interval to eliminate second spike
  9. Make sure PSG (Post synaptic conductance) module is on
  10. Run PRC for a few minutes, look to see if synaptic inputs advance spikes by eye. If not, increase PSGScalar (which is the size of synaptic inputs
  11. Open matlab from prompt by typing "matlab"
  12. Analyze data in Matlab:
    1. cd to directory where data was recorded
    2. load data into Matlab using: Data=DataLoad('yourfilename');
    3. plot PRC of data: plotPRC(Data,0.1,[],0);  %plotPRC(Data,SpikeRateInSeconds,SigTimes,ReversalPotentialInmV)
  13. If PRC looks like noise (i.e. no spike time advance dependency on phase) crank up the synaptic input strength (PSGScalar in PSG module) and take another data set
  14. If PRC is pinned to maximal advance on right hand side, decrease synaptic strength (PSGScalar in PSG module)
  15. If it looks good, then save the database file and do whatever experimental adjustments to the cell you desire

Noisy Networks

This experiment consists of 4 recordings for each complete experiment.

  1. Driving neuron with Poisson drive and measuring spike times
    1. Patch neuron
    2. Load settings file "NoisyNet_Ch0_C1.set"
    3. Read in Poisson Event time file for playback "/root/rtxi/settings/PRN_500_150000_C1.evt"
    4. Set event recording file to "/root/rtxi/Data/Date_NN_C1_CXXX_1.evt"
    5. Set filename in Data record file to the same name as the event recorder name with ".nds" as the name extension
    6. In synch module: remove ", 15" which is the event recorder from the list
    7. Un-pause the synch module to turn spike rate controller on and Poisson drive. When cell firing rate stabilizes pause the synch funciton
    8. Add Date_NN_C1_CXXX_1.evt to lab notebook with note:

      Cell 1, poisson Drive

      Virtual cell being driven with event file:
      PRN_500_150000_C1.evt
      500 spks/sec 5 minutes long
      PSG:
      Rise=0.0015
      Fall=0.005
      Amp=2
      Delay=0

    9. Add the event record module back to synch by adding ", 15" to the list
    10. Un-pause the synch function to start everything running.
    11. Un-pause the Data Recorder, record for a little while and re-pause
    12. The percent complete of the Poisson drive shows in the event playback module in the 2nd to top box. When this reaches 100, This takes 5 minutes.
    13. pause everything with the synch function.
    14. You will get a notification box asking if you want to close the event record file, select "Close"
  2. Driving neuron with Poisson drive + synaptic input from spike times from first cell and measuring spike times
    1. Load settings file "NoisyNet_Ch0_C2.evt"
    2. Read in Poisson Event time file for playback "/root/rtxi/settings/PRN_500_150000_C2.evt"
    3. Read in the event times you just recorded for playback "/root/rtxi/Data/Date_NN_C1_CXXX_1.evt"
    4. Set event recording file to "/root/rtxi/Data/Date_NN_C2_CXXX_1.evt"
    5. Set filename in Data record file to the same name as the event recorder name with ".nds" as the name extension
    6. Add Date_NN_C2_CXXX_1.evt to lab notebook with note:

      Cell 2, poisson Drive

      Virtual cell being driven with event file:
      PRN_500_150000_C2.evt 500 spks/sec 5 minutes long
      Date_NN_C1_CXXX_1.evt, event file from same cell recorded earlier

      PSG for poisson input:
      Rise=0.0015
      Fall=0.005
      Amp=2
      Delay=0

      PSG for Cell1 input:
      Rise=0.0015
      Fall=0.005
      Amp=2
      Delay=0

    7. Repeat steps 6-13 from the above experiment
    8. In Matlab, run the function "SpikeDiffHist('/root/rtxi/Data/Date_NN_C1_CXXX_1.evt','/root/rtxi/Data/Date_NN_C2_CXXX_1.evt');
    9. If correlation is very strong or is very weak, repeat last experiment and this experiment with new synatpic strengths to find a strength with weak but visible crosscorrelation between the two cells
  3. Driving neuron with Poisson drive + synaptic input from spike times with strong synaptic strength from first cell and measuring spike times
    1. This is exactly the same experiment as the last but with slightly stronger synapses for PSG 21 module. Increase it by 2 to see stronger correlation between C1 and C2, this is so that when we do C3, the correlation may be strong enough with the common cell that their correlations will look strong too.
    2. Set event recording file to "/root/rtxi/Data/Date_NN_C2_CXXX_2.evt"
    3. Set filename in Data record file to the same name as the event recorder name with ".nds" as the name extension
    4. Add Date_NN_C2_CXXX_2.evt to lab notebook with note:

      Cell 2, poisson Drive

      Virtual cell being driven with event file:
      PRN_500_150000_C2.evt 500 spks/sec 5 minutes long
      Date_NN_C1_CXXX_1.evt, event file from same cell recorded earlier

      PSG for poisson input:
      Rise=0.0015
      Fall=0.005
      Amp=2
      Delay=0

      PSG for Cell1 input:
      Rise=0.0015
      Fall=0.005
      Amp=4
      Delay=0

    5. repeat steps 6-13 from first experiment.
  4. Driving neuron with Poisson drive + synaptic input from spike times from first cell plus a delay and measuring spike times
    1. Read in Poisson Event time file for playback "/root/rtxi/settings/PRN_500_150000_C3.evt"
    2. Read in the event times you just recorded for playback "/root/rtxi/Data/Date_NN_C1_CXXX_1.evt"
    3. Set event recording file to "/root/rtxi/Data/Date_NN_C3_CXXX_1.evt"
    4. Set filename in Data record file to the same name as the event recorder name with ".nds" as the name extension
    5. Set synatpic delay in PSG module to 2
    6. Load settings file "NoisyNet_Ch0_C2.evt" and set all files as above
    7. Add Date_NN_C2_CXXX_1.evt to lab notebook with note:

      Cell 2, poisson Drive

      Virtual cell being driven with event file:
      PRN_500_150000_C2.evt 500 spks/sec 5 minutes long
      Date_NN_C1_CXXX_1.evt, event file from same cell recorded earlier

      PSG for poisson input:
      Rise=0.0015
      Fall=0.005
      Amp=2
      Delay=0

      PSG for Cell1 input:
      Rise=0.0015
      Fall=0.005
      Amp=4
      Delay=0.01

    8. Repeat steps 6-13 from first experiment
    9. Contact Duane and tell him you have a complete set of experiments.

Decrease synaptic strength and incrase rate for noise drive to neurons. Find spike time variance/rate/strength relationship and optimize for realistic synapse strength with reasonable variability
Decrease time constant for synapse.
Find Neuron's synaptic strength required to get significant correlation between pre & postsynaptic neuron.
Use this first strength for synapses with direct connection experiment.
Increase synaptic strength and repeat experiment with cell 2 for anlayis on common input experiment.
Repeat with second post-synaptic neuron with longer synaptic delay. Set that delay equal to the lag with the peak response in between the first pair. It was 5msec for the original synapse, changing synaptic shape may change this delay.
Look at Kelvin's last experiments to find details of settings. Files for generating noisy inputs and correlated noisy input files are in Kelvin's directory.

Record voltage traces Try different amounts of correlated noise between cell 2 and cell 3 drive and cell 1.

About PID

On July 20, 2009, the PID module on rtxi was modified. This change is taking into account that the dt for the error integration is the isi instead of the sampling time. Be careful to analyze data obtained using that module before that day.

Personal tools
Namespaces

Variants
Actions
Navigation
Toolbox