Skip to main content

Table 3 Performance evaluation under Study II (Setting 1), (G=5000, 500 DE gene) #

From: Confident difference criterion: a new Bayesian differentially expressed gene selection algorithm with applications

Method

CDE

CCDE

CCEE

FNR

FPR

FDR

FNDR

twocri. (γ 0=0.4)

785.1(8.3)

465.6(2.5)

4180.5(7.6)

0.069(0.005)

0.071(0.002)

0.407(0.006)

0.008 (0.001)

(γ 0=0.6)

509.4(6.8)

421.3(4.6)

4411.9(4.1)

0.157(0.009)

0.019(0.001)

0.173(0.007)

0.018(0.001)

(ζ=0.05)

344.2(8.2)

328.8(6.3)

4484.6(2.6)

0.342(0.013)

0.003(0.001)

0.044(0.007)

0.037(0.001)

edgeR 1(ζ=0.05)

289.7(17.6)

278.1(16.5)

4488.3(3.6)

0.444(0.033)

0.003(0.001)

0.040(0.011)

0.047(0.003)

edgeR 2(ζ=0.05)

290.6(18.1)

276.4(16.8)

4485.8(3.8)

0.447(0.034)

0.003(0.001)

0.049(0.012)

0.047(0.003)

DESeq(ζ=0.05)

297.2(21.3)

265.9(18.4)

4468.7(5.6)

0.468(0.037)

0.007(0.001)

0.105(0.016)

0.050(0.002)

BaySeq(ζ=0.05)

203.1(22.8)

203.0(22.8)

4499.9(0.2)

0.594(0.046)

0.000(0.000)

0.000(0.001)

0.062(0.004)

NBPSeq(ζ=0.05)

248.3(20.5)

239.8(19.3)

4491.5(4.0)

0.520(0.039)

0.002(0.001)

0.034(0.015)

0.055(0.004)

EBSeq(ζ=0.05)

303.7(18.8)

257.7(14.9)

4454.0(6.8)

0.485(0.030)

0.010(0.002)

0.151(0.017)

0.052(0.003)

NOISeq(ζ=0.05)

303.1(19.1)

294.4(17.6)

4491.3(3.2)

0.411(0.035)

0.002(0.001)

0.028(0.010)

0.044(0.004)

SAMSeq(ζ=0.05)

134.2(45.2)

126.1(43.2)

4491.9(3.4)

0.748(0.086)

0.002(0.001)

0.061(0.022)

0.077(0.008)

TSPM(ζ=0.05)

85.4 (19.2)

58.7 (15.4)

4473.3(6.5)

0.883(0.031)

0.006(0.001)

0.316(0.056)

0.090(0.003)

  1. #Empirical estimates of the standard deviation were reported in the parentheses
  2. edgeR 1estimates the common dispersion parameter for all tags; edgeR 2 estimates the tag-wise dispersion parameters
  3. ζdenotes the FDR