Skip to main content

Table 2 The AUCs of edgeR-robust, DESeq2, limma-voom, ELMSeq and rSeqRobust based on log-normally distributed data

From: Joint between-sample normalization and differential expression detection through 0-regularized regression

DE (%)Up (%)edgeRDESeq2voomELMSeqrSeqRobust
1500.97340.97360.97860.97170.9757
  (0.0091)(0.009)(0.007)(0.0065)(0.0064)
1750.9540.95310.97110.93430.935
  (0.0113)(0.0141)(0.0086)(0.0153)(0.018)
11000.95250.95310.96330.94760.9594
  (0.0144)(0.0137)(0.0108)(0.0151)(0.0139)
10500.9580.96230.96680.95730.9627
  (0.0079)(0.0069)(0.0057)(0.0069)(0.0067)
10750.97070.96320.97490.9640.9668
  (0.0057)(0.0045)(0.004)(0.0061)(0.0057)
101000.94030.92720.96050.940.9435
  (0.0107)(0.0142)(0.0077)(0.0128)(0.0128)
30500.96890.96960.9740.96650.9678
  (0.0056)(0.0048)(0.005)(0.0053)(0.0052)
30750.93180.92650.94580.95640.9655
  (0.0113)(0.0116)(0.0096)(0.0078)(0.0059)
301000.87710.86930.87530.93720.9566
  (0.0153)(0.0091)(0.0145)(0.0149)(0.0086)
50500.94660.9540.94250.95570.957
  (0.0092)(0.0059)(0.0087)(0.0071)(0.0065)
50750.90990.9060.91450.94010.9566
  (0.0167)(0.0123)(0.0178)(0.0135)(0.0076)
501000.70830.72360.71970.8790.9576
  (0.022)(0.0291)(0.0242)(0.0195)(0.0071)
70500.9670.96550.96520.96550.969
  (0.0039)(0.0034)(0.0036)(0.0031)(0.0021)
70750.85690.83510.85640.90890.9692
  (0.0193)(0.0161)(0.0194)(0.0118)(0.0045)
701000.45360.52120.48930.47860.9638
  (0.0344)(0.0296)(0.018)(0.037)(0.0082)
90500.9530.95380.95130.95610.9512
  (0.0064)(0.0064)(0.0081)(0.0042)(0.0049)
90750.72030.69180.72560.69060.9584
  (0.0239)(0.0177)(0.0323)(0.0167)(0.0084)
901000.25680.5060.25660.35160.8276
  (0.0257)(0.0265)(0.0278)(0.0345)(0.0426)
  1. The sample size is n=20. The variance of the normal distribution is \(\sigma _{i}^{2}=0.01\). The table shows the percent of DE genes (DE %), percent of up-regulated genes among all the DE genes (Up %), and the mean AUCs (standard errors in parentheses) for all five methods with 10 simulated replicates. The highest AUC value is shown in bold