For the ANOVA, the Genotype term tested for a difference between wild-type and the 35S MYB lines while the impact of the two independent lines per 35S MYB transgene were simultaneously tested by the nested factor, Transgene Genotype).
By fitting ordinal exposure variables as continuous terms, tests for dose-response trends in risks of ovarian cancer were carried out.
Terms are tested for matches in descending order of length, so the longest possible matching term is always replaced with the appropriate CTERM or GTERM token.
To ensure independence of error terms we tested for autocorrelation in the residuals from all regression models using the Durbin-Watson statistic [17].
Finally, theoretically relevant interaction terms were tested for significance.
Plausible interaction terms were tested for statistical significance.
For downstream analysis, GO [ 70] biological process terms were tested for enrichment.
All interaction terms were tested for colinearity before inclusion in the models.
The z ij terms were tested for inclusion in models using a forward selection procedure for entering z ij.
Differentially expressed genes were mapped onto Gene Ontology (GO), and GO terms were tested for overrepresentation using the hypergeometric distribution [ 26].
