Skip to main content
Figure 1 | BMC Bioinformatics

Figure 1

From: Towards precise classification of cancers based on robust gene functional expression profiles

Figure 1

Training classification rules for four cancer types based on functional expression profiles of 114 modules. A – Decision tree trained with the NCI60 FEP median measure. The internal nodes of the tree are denoted with the functional modules from Gene Ontology. The leaf nodes give the classification results for the cancer types. The numbers in the leaf nodes are the total number of samples contained over the number of the incorrectly predicted samples. B – Functional expression profiles of the three identified modules. For the identified GO modules from decision analysis, their functional expression profiles are demonstrated with a colouring spectrum of their medians. Each GO module corresponds to a row, and the column denotes the functional expression for each cell line. At the top are names of cell lines (renal cancer (RE), colon cancer (CO), leukaemia (LE), melanoma (ME)). Samples with a missing value or the null value are coded with black colour, a positive with red colour and a negative with green colour. C – numbers of genes annotated and differentially expressed in the three identified modules.

Back to article page