This is the readme for the model associated with the paper:

Gerkin RC, Lau PM, Nauen DW, Wang YT, Bi GQ (2007) 
Modular Competition Driven by NMDA Receptor Subtypes in
Spike-Timing-Dependent Plasticity 
J Neurophysiol. (epub ahead of print)

A native igor program, a text version for those without igor, and a
rich text format version of the (below) instructions is supplied in
this archive.

The parameters are set to the values used in the paper.  (Gerkin et
al., J. Neurophys., 2007) They can be restored to these values by
pressing "Reset".  The simulation can be started by pressing "Go".
The resulting evolution of "W" from the model will be displayed in the
graph titled "Results".  Each time "Go" is pressed the trace in the
"Results" graph will be overwritten by the values obtained in the
latest simulation.  The code can be examined in the experiment's
Procedure Window.

Parameters (most are described in the Methods section of the paper):  

duration: duration of the simulation in ms.  
frequency: frequency of the stimuli (e.g. a doublet or triplet is one
stimulus) in Hz.
pulse_width: "effective width" of the doublets, i.e. how long do they
act on the system?
ba_On: Whether or not the AB doublet is "on" (0 or 1).   
ab_On: Whether or not the BA doublet is "on" (0 or 1).  
timing: Interval between doublets (time of BA - time of AB).  +10
would be the ABA triplet, and -10 would be the BAB triplet.
dt: The integration timescale for the equations.  Smaller values are
more accurate, but take more time to simulate.  Values much above 0.01
can yield errors.
p_off: The rate at which the potentiation process P decays back to 0
after being activated.
d_off: The rate at which the depression process D decays back to 0
after being activated.
v_off: The rate at which the veto process V decays back to 0 after
being activated.
lambda: The effect of the veto process on the depression process.  
p_steep: The steepness of the relationship between the synaptic
modification W and the potentiation process P.
d_steep: The steepness of the relationship between the synaptic
modification W and the depression process D.
w_off: The rate at which the synaptic modification W decays back to 0
after being activated.
nr2a: The relative amount of NR2A (linked to P) that is available.  
nr2b: The relative amount of NR2B (linked to D and V) that is available.  

Figure 5B was generated by simulating all combinations of ab_on,
ba_on, nr2a, and nr2b (each set to 0 or 1), with timing equal to
either -10 or +10.  Figure 5C was generated by simulating all
combinations of ab_on, ba_on, with nr2a=0 and nr2b=2/3 (according to
the experimental effects of 100 nM NVP-AAM077, as described in Figure
1), and timing set as above.