Table 2. Best fitting scores for the total ODE model described in Fig. 3, according to the saturation and the dose-dependent hypotheses.
This table depicts the fitting score of each metabolic compounds according to the parameter estimation procedure detailed in Material and Methods. The p-value is a score reported by the PottersWheel toolbox in order to evaluate the goodness of the best-fitting set of parameters by taking into account the number of datapoints and the quality of the inferred sets of parameters. Local and global Chi2 depict more precisely the quality of such sets of parameters. A local score is considered to be valid (✓) when it is lower than 1 and when its value is located between the errors bars computed by The PottersWheel toolbox at t = 24 h. A global score is valid when all local scores are valid. In other cases, the scores are considered as invalid ✗ and the corresponding hypotheses are rejected.
| µMol | Best Chi2 score per 1,000 fit sequence | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MeIQx | MeIQx | MeIQx-N2-Gl | MeIQx-N2-SO3- | oxo-MeIQx | C-Hydroxy-MeIQx | HON-MeIQx-N2-Gl | HONH-MeIQx | Potential-Genotoxic-Compound | Local Chi2 score | Global Chi2 score | Best p-value | |
| Saturation hypothesis 1. Does the pathway saturation explain the data? | ||||||||||||
| The CYP1A2 and UGT distribution coefficients α and β are considered as global parameters, which | ||||||||||||
| are independent of the initial MeIQx concentration and have the same value for the 3 initial doses of MeIQx | ||||||||||||
| Fixed α and β | 1 | ✓ (0.01) | ✓ (0.82) | ✓ (0.00) | ✓ (0.00) | ✓ (0.00) | ✓ (0.01) | ✓ (0.00) | ✓ (0.01) | ✓ (0.854) | ✗ 20.9622 | ✗ 0.893335 |
| 10 | ✓ (0.01) | ✗ (12.98) | ✓ (0.00) | ✓ (0.00) | ✗ (1.88) | ✓ (0.23) | ✓ (0.00) | ✓ (0.39) | ✗ (15.50) | |||
| 50 | ✓ (0.04) | ✓ (0.35) | ✓ (0.00) | ✓ (0.00) | ✗ (2.99) | ✓ (0.02) | ✓ (0.00) | ✗ (1.21) | ✗ (4.610) | |||
| Dose-dependent hypothesis. Can we model the dynamic | ||||||||||||
| with a function directly dependent of the substrate concentration? | ||||||||||||
| The fitting experiments are performed by considering that the distribution coefficients | ||||||||||||
| α and β depend on the concentration of MeIQx | ||||||||||||
| α and β dynamically evolve with MeIQx concentration in function of time. | ||||||||||||
| These parameters are modeled by sigmoid functions which depend on the MeIQx concentration. | ||||||||||||
| α and β dynamical | 1 | ✓ (0.84) | ✓ (0.00) | ✓ (0.01) | ✓ (0.05) | ✓ (0.02) | ✓ (0.03) | ✓ (0.00) | ✓ (0.04) | ✓ (0.987) | ✓ 2.57838 | ✓ 1 |
| 10 | ✓ (0.58) | ✓ (0.03) | ✓ (0.35) | ✓ (0.30) | ✓ (0.02) | ✓ (0.07) | ✓ (0.00) | ✓ (0.01) | ✓ (1.34) | |||
| 50 | ✓ (0.04) | ✓ (0.00) | ✓ (0.00) | ✓ (0.00) | ✓ (0.00) | ✓ (0.01) | ✓ (0.00) | ✓ (0.19) | ✗ (0.248) | |||