Figure 2.

Workflow. The workflow of a set-level learning experiment conducted multiple times with varying alternatives in the numbered steps. For compatibility with the learned classifier, testing fold samples are also reformulated to the set level. The reformulation is done using gene sets selected in Step 3 and aggregation algorithm used in Step 4. The diagram abstracts from this operation.

Holec et al. BMC Bioinformatics 2012 13(Suppl 10):S15   doi:10.1186/1471-2105-13-S10-S15