Skip to main content

Table 3 Result of the iDEGES approach does not necessarily convergent

From: TCC: an R package for comparing tag count data with robust normalization strategies

P G1 50% 70% 90% Total
(a) iDEGES/TbT     
convergent 69 70 83 222
cyclic 31 30 17 78
(b) iDEGES/edgeR     
convergent 86 86 79 251
cyclic 14 14 21 49
(c) iDEGES/TDT     
convergent 82 81 84 247
cyclic 18 19 16 53
(d) iDEGES/DESeq     
convergent 99 97 100 296
cyclic 1 3 0 4
  1. The numbers of convergent and non-convergent (cyclic) results of 100 trials under PDEG = 20% are shown: (a) iDEGES/TbT, (b) iDEGES/edgeR, (c) iDEGES/TDT, and (d) iDEGES/DESeq. We defined the trial as ‘convergent’ if potential DEGs estimated in the (NC + 1)th iteration was the same as those in the (NC)th iteration and the number of iteration n required for obtaining the convergent result as NC (NC > = 1). We defined the trial as ‘cyclic’ if potential DEGs estimated in the (i + NP)th iteration were the same as those in the ith iteration and the cycle as NP (NP > = 2).