Fig. 2.

The workflow of ContextTRAP. Time-series gene expression data is given to TRAP and keywords representing the context are given to BEST as input. ContextTRAP incorporates TRAP and BEST by integrating the analysis results from both tools. The entity score derived from BEST with keywords is converted to Context Score (CS), which measures the relevance of pathway. For that, the concept of posterior probability and Bayes’ rule are used. The p-value, the result of TRAP, is transformed and normalized to Impact Score (IS) measuring the significance of pathway. Then, Discovery rate is automatically determined as a weight for integration of two scores, depending on the results of TRAP and BEST. Finally, the integrated score is used to prioritize pathways in terms of both significance and relevance