Auditory PWM task data and code

The data underlying the figures in Akrami et al. Nature (2018) is publicly available and can be found here.

The code at the link below will run under the Bcontrol training system. Using this code will automatically train a rat to perform the “Auditory Parametric Working Memory” (aPWM) task described in Akrami et al. Nature (2018).

Download from here or here Automated Automated Parametric Working Memory Training

After downloading, unzip the file, and you will find the following directory and four files.

  • The directory “@AthenaDelayComp” contains the code for a protocol that we use to initialize/shape/train rats on different stages of the aPWM task; with this protocol they
    • first learn to associate LEDs in the right and left nose ports with water rewards (Stage0_NP)
    • then learn to associate poking in the center poke with water rewards delivered from the side poke (Stage0_NP)
    • then learn to increase the duration of poking in the center poke, up to 10 seconds (Stage1_NPGrowing)
    • then learn the delayed comparison rule: first_stimulus > second_stimulus => right, first_stimulus < second_stimulus => left (IntroducingStimuli & DelayedReward)
  • The file “settings_@AthenaDelayComp_Athena_Stage0_NP.mat”  contains the settings to initialize training on the first phase of the training for a naive rat. Note that they’re set for a rat called “W100” running under experimenter “Athena”, starting on Sep 24, 2012– you should of course change that to suit your own needs.
  • *settings_@AthenaDelayComp_Athena_Stage1_NPGrowing.mat*
    • trains the animal to nose poke and wait for the Go cue. Nose poke duration grows with a certain rate (for instance 1% increase/trial).
  • *settings_@AthenaDelayComp_Athena_IntroducingStimuli.mat*
    • auditory stimuli are presented, and reward is guaranteed to be delivered based on the rule set (s_a > s_b -> right; s_a < s_b -> left)
  • *settings_@AthenaDelayComp_Athena_DelayedReward.mat*
    • two final stages of the training: “delayed reward” and then “no reward” on the error trials

To train a rat, set it on the “Stage0_NP” of the protocol. Usually a week suffices. The settings will automatically update as they need from one day to the next.

After that, start the rat switch the protocol to “Stage1_NPGrowing”. Wait ~1 months to reach to the NP duration of 8-10second.

After that, start the rat switch the protocol to “IntroducingStimuli”. After 3-5 day switch to “DelayedReward”. Wait ~2-3 months!

Remember: although training will run in an automated manner, it is always good to monitor your rats!  Individuals sometimes need adjustments.