Table 1. Summary statistics for the linear modelling of uORF repressiveness and CDS TE using various sequence feature sets.
| Parameter modelled | Sequence feature set | Pearson r | P value | Fold change | PRESS | RESS | n |
|---|---|---|---|---|---|---|---|
| uORF repressiveness | uORF | 0.2306 | 1.64 × 10−10 | 4.11 | 3,682 | 3,643 | 750 |
| uORF+5′ leader | 0.2371 | 4.79 × 10−11 | 4.45 | 3,689 | 3,632 | 750 | |
| uORF+CDS | 0.288 | 8.71 × 10−16 | 5.609 | 3,602 | 3,529 | 750 | |
| uORF+CDS+5′ leader | 0.2922 | 3.14 × 10−16 | 5.979 | 3,610 | 3,519 | 750 | |
| CDS TE | CDS | 0.3468 | 1.29 × 10−22 | 3.175 | 1,103 | 1,089 | 750 |
| CDS+uORF | 0.421 | 1.42 × 10−33 | 4.227 | 1,047 | 1,019 | 750 | |
| CDS+5′ leader | 0.4148 | 1.50 × 10−32 | 4.099 | 1,043 | 1,025 | 750 | |
| CDS+5′ leader+uORF | 0.4421 | 3.05 × 10−37 | 4.337 | 1,029 | 996 | 750 |
CDS, coding sequence; TE, translation efficiency
Summary statistics includes the linear correlation and prediction errors (using the predicted residual sum of squares or PRESS statistic), as well as the fold change in uORF repressiveness observed over a 95% range in the combined scores.