Item | Redundancy r^2 | VIF | Step | Multiple r | Multiple r^2 | r^2 increase | F | P | Partial r |
---|
Lmi_subfam | 0.668 | 3.01 | Â | Â | Â | Â | Â | Â | Â |
Lmi_peak_copies | 0.262 | 1.36 | Â | Â | Â | Â | Â | Â | Â |
Lmi_TSI | 0.331 | 1.50 | Â | Â | Â | Â | Â | Â | Â |
Lmi_RPS | 0.236 | 1.31 | Â | Â | Â | Â | Â | Â | Â |
Ode_subfam | 0.118 | 1.13 | 1 | 0.751 | 0.564 | 0.564 | 15.54 | 0.0020 | 0.845 |
Ode_peak_copies | 0.067 | 1.07 | 2 | 0.906 | 0.822 | 0.257 | 15.89 | 0.0021 | 0.758 |
Ode_TSI | 0.172 | 1.21 | 3 | 0.922 | 0.850 | 0.028 | 1.89 | 0.1995 | 0.398 |
Ode_RPS | 0.565 | 2.30 | Â | Â | Â | Â | Â | Â | Â |
- In each species, the independent variables employed were the number of subfamilies (subfam), the number of repeat units included in amplification peaks (peak_copies), the tandem structure index (TSI) and the homogenization index (RPS). Note that only three independent variables entered in the model, all of them corresponding to Ode, and only two (Ode_subfam and Ode_peak_copies) were associated with significant increases in explained variance in CTR (56.4% and 25.7%, respectively. The multiple correlation coefficients were 0.652 (SE= 0.13) and 0.466 (SE= 0.127), respectively. The Shapiro-Wilks test showed that the standardized residuals of this regression fitted a normal distribution (W= 0.966, P= 0.821). VIF= Variance inflation factors. Redundancy r^2 was performed between each independent item and the seven remaining, in order to calculate VIF as 1/(1-r^2)