Skip to main content

Table 1 Identifiability ranking and results of multi-start estimation and variance analysis

From: A simple work flow for biologically inspired model reduction - application to early JAK-STAT signaling

Model Top ten parameters of the identifiability ranking χ 2 AICc v(pi)
M 0 kf7, kf9, kd11, kd9, kf24, kf13, kd5, kd13, kd7, kf1 none none v(kf7) ≥ 0.63 > v ¯
M 1 kd5, kf9, kd11, kf13, kf1, kf25, kd13, kd9, kd1, kd24 8.7 · 10-6 77 v(kd5) ≥ 0.6 > v ¯
M 2 kd1, kf25, kf1, kf10, kd3, kf12, kf13, k24, kd13, k11 1.0 · 10-4 57 v(kd1) ≥ 0.59 > v ¯
M 3 kf25, kf10, kf12, k24, kd3, kf13, kd13, k11, k9, kf3 9.5 · 10-5 49 v(kf25) ≥ 0.75 > v ¯
M 4 k5, kf10, kd3, k9, kf3, kd8, kf4, k11, kf8, kd2 2.7 · 10-3 31 v(k5) ≥ 0.47 > v ¯
M 5 k11, kf10, kd2, kf3, kd3, kd8, k9, kf4, kf8, kf2 7.6 · 10-1 28 v(k11) ≥ 0.18 > v ¯
M 6 kf10, kd2, kf3, kd3, kd8, k9, kf4, kf8, kf2, k5new 7.6 · 10-1 25 v(p i ) ≤0.0091 < v ¯
  1. The first parameter in each ranking is the least identifiable one. No χ2 and AICc values are given for M0, since the model parameters are not estimated but taken from the literature [36].