Supplementary Table 3 Lower and upper bounds of parameters uniform distributions used to generate initial parameters for fitting, and hard limits enforced by pattern search during the optimisation process
Parameter | Symbol | Initial parameter distribution | Hard limits enforced by optimizer | ||
|---|---|---|---|---|---|
Lower bound | Upper bound | Lower bound | Upper bound | ||
I to E weight | \(w_{IE}\) | 0 | 10 | 0 | 30 |
E to I weight | \(w_{EI}\) | 0 | 10 | 0 | 30 |
E to E weight | \(w_{EE}\) | 0 | 10 | 0 | 30 |
Sigmoid steepness parameter | β | 0 | 10 | 0 | 30 |
Time constant (s) | τ | 0 | 0.3 | 0 | 0.5 |
Constant input to E | \(\theta_{E}\) | −2 | 10 | −30 | 30 |
Constant input to I | \(\theta_{I}\) | −10 | 2 | −30 | 30 |
Noise standard deviation | ζ | 0 | 0.1 | 0 | 0.3 |
Stimulation magnitude | δE | 0 | 0.02 | 0 | 0.1 |
Stimulation delay (ms) | \(\Delta t_{\text{stim}} \) | 0 | 250 | 0 | 500 |