Skip to main content

Table 2 Improvement in biological coherence after iterative clustering of dataset II

From: Microarray data mining: A novel optimization-based approach to uncover biologically coherent structures

 

Proportion of Genes (%)

 

p-value <= 10-4

p-value <= 10-5

p-value <= 10-4

p-value <= 10-5

 

(Uncorrected)

(Bonferroni-Corrected)

Initial Iteration

79.4

66.1

57.2

52.2

Final Iteration

86.2

83.9

73.1

68.9

  1. Shown are the percentages of genes in clusters in which one or more subsets of the genes in the cluster exhibit a statistically non-random membership in a common biological function, as defined by Gene Ontology (GO) classification, either at a significance level of 10-4 or 10-5, either uncorrected (left) or Bonferroni corrected for multiple hypothesis testing (right). The values are for the clusters obtained following initial EP_GOS_Clust clustering or after six round of the iterative algorithm described in Methods.