Spiking neuron model stimulated through an auditory transduction chain.

Author

Alexander Wolf

Jan Benda

ReceptorModel simulates the auditory transduction chain of Locusta migratoria. It is designed to work with RELACS in the dry mode (relacs -3) so new repros can be tested for different settings. The plugin is based on a biophysical model that describes the transformation of the sound stimulus in a chain of three distinct steps: 1. the tympanum; 2. mechano-electrical transduction; 3. spike generator.

For each of the three steps, the plugin provides several options for the user to adapt to his/her needs and the hardware restrictions. Additionally, the user can chose between between several integration algorithms. The following options can be selected (in order of increasing complexity):

1. Tympanum

• None: No modification of the stimulus
• Scaling: Adjusts the stimulus strength according to tuning curve
• Oscillator: Model as damped harmonic oscillator

1. Mechano-Electrical Transduction:

• None: No modification of the tympanum output
• Linear: Absolut value of the tympanum output
• Linear saturated: Absolut value up to saturation point
• Box: Two Heavyside functions
• Square: Square of the tympanum output
• Square saturated: Square value up to saturation point
• Linear Boltzman: Superposition of two Boltzman functions with linear dependency in exponent
• Square Boltzman: One Boltzman function with quadratic dependency in exponent

1. Spike Generators:

• Stimulus: No modification of previous function
• FitzhughNagumo: Simplified abstract neuron model with 2 differential equations.
• MorrisLecar: Simplified neuron model based on ion channels with 2 differential equations.
• WangAdapt: Neuron Model with adaptation currents with 5 differential equations.

1. Integrators:

• Euler
• Midpoint
• Runge-Kutta 4th order

Options

The following options are supported (brief description of each option with default values and data type):

• Transduction chain
• tymp=Scaling: Tympanum model (string)
• freq=5kHz: Eigenfrequency (number)
• tdec=0.154ms: Decay constant (number)
• nl=Square saturated: Static nonlinearity (string)
• Spike generator
• spikemodel=Stimulus: Spike model (string)
• noise=0: Standard deviation of current noise (number)
• deltat=0.005ms: Delta t (number)
• integrator=Euler: Method of integration (string)
• Square = ax^2+imin, a=(imax-imin)/cut^2
• Square saturated = imax, for |x|>=cut
• Linear = b|x|+imin, b=(imax-imin)/cut
• Linear saturated = imax, for |x|>=cut
• Box = imin, for |x|<cut, = imax else
• None = ax, a=(imax-imin)/cut
• imax=60muA/cm^2: Maximum current (number)
• imin=0muA/cm^2: Minimum current (zero point current) (number)
• cut=0.02mPa: Amplitude of tympanum where imax is reached (number)
• Boltzmann, (imax/(1-f_0))*(1/(1+exp[-slope*(x-x0)])+1/(1+exp[slope*(x+x0)])-f_0)+imin
• slope=3000mPa^-1: Slope of Boltzmann (number)
• matchslope=true: Set slope of Boltzmann to match square (boolean)
• x0=0.01mPa: 1/2 of Imax-Imin is reached (number)
• Boltzmann, 2(imax-imin)(1/(1+exp[-slope2*x^2])-1/2)+imin
• slope2=6e+05mPa^-2: Slope of square Boltzmann (number)
• matchslope2=true: Set slope of square Boltzmann to match square (boolean)
• Many more options for spike generators.

Version

1.1 (Jan 15, 2006)