Table 1. Linear model results for secreted and nuclear proteins at day 5.
| SECRETOME | NUCLEAR | ||||||
| Polysomal RNA | Polysomal RNA | ||||||
| logFC |
|
|
|
logFC |
|
|
|
†
|
- | - |
|
|
- | - |
|
†
|
miR-130a† | - |
|
|
miR-185*† | - |
|
†
|
miR-130b† | - |
|
|
miR-20b*† | - | 0.175 |
†
|
miR-130b† | miR-558† |
|
|
miR-16-2*† | miR-185*† |
|
Results for applying linear models to the data at day 
secreted and nuclear proteins. Both RNA fractions are considered. For each subtable (e.g. secretome-polysomal) the first row shows the results for a linear model without considering microRNA effect (the standard model: 
vs. 
). The 2nd and 3rd row represent the values for univariate models, including the effect of only one miRNA. We selected the two most significant miRNAs. The last row shows the (multivariate) best model as determined by the BIC value. In several cases the best model is not multivariate, especially since BIC penalizes the number of parameters.
†
means a significance level of 
.