Skip to main content
Figure 5 | Genome Biology

Figure 5

From: Strong-association-rule mining for large-scale gene-expression data analysis: a case study on human SAGE data

Figure 5

Rules generated. (a) The 68 rules generated by ac-miner on a 'max minus 25%' boolean matrix using a 5% frequency threshold. The rules were translated into a color-coded Excel file. For the color-coding conventions used, see Figure 2. (b) The first three rules generated. The first rule reads: 'When the gene encoding the ribosomal protein S24 (identified by the tag no. 150) is overexpressed, then the gene encoding the cytochrome c oxidase subunit IV (identified by tag 255) is also overexpressed'. This rule is true in four different libraries. The second rule associates three tags identifying the ribosomal proteins S5, L19 and S18 (true in five libraries). The third rule associates five tags identifying the ribosomal proteins L21, S5, S23, L19 and S18 (true in four libraries).

Back to article page