Skip to main content

Table 1 Performance comparison for three methods

From: A stochastic model for identifying differential gene pair co-expression patterns in prostate cancer progression

Data type Statistics Methods
   SIG method ECF method DAGA method
HS vs. HR Identified pairs 428,582 2,101,425 5,212,394
  RS 1.49% 100% 91.07%
  P-value of RS 0.011 1 0.93
  PS_Avg 1.74 0 0.02
  P-value of PS_Avg 0.032 1 0.97
  TFR 1.39% 0.49% 0.26%
  PES_Avg 3.56 2.56 2.02
cancer vs. healthy Identified pairs 303,992 14,228,478 32,569,223
  RS 1.40% 100% 96.21%
  P-value of RS 0.013 1 0.87
  PS_Avg 1.65 0 0.01
  P-value of PS_Avg 0.036 1 0.99
  TFR 0.92% 0.39% 0.25%
  PES_Avg 3.32 2.36 1.86
  1. RS: Re-identification Scale, overlap between the pairs identified before and after changing disease progression relation; PS: Progression Score, ratio of the nominal P-value of a pair before changing disease progression relation to that of a pair after changing disease progression relation; TPR: True Positive Rate, the proportion of functionally relevant pairs in all characterized pairs; PES: Pathway Enrichment Score, the degree to which differentially co-expressed gene pairs are overrepresented at a pathway. The 'Avg' refers to the average value for an item.