Skip to main content

Table 1 Efficiencies of different methods in identifying genes differentially expressed among four groups each with 6 replicates in 30 simulated datasets

From: Ranking analysis of F-statistics for microarray data

   NGCS ENFP TNFP Difference between ENFP and TNFP
Method FDR Mean (SD) Min Max Mean (SD) Min Max Mean (SD) Min Max | d ¯ | MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGaciGaaiaabeqaaeqabiWaaaGcbaWaaqWaaeaacuWGKbazgaqeaaGaay5bSlaawIa7aaaa@3060@ Var (d) C(d ≥ 0)
B procedure λ = 0.05 59.6 (6.6) 46 73 3.0 (0.3) 2 4 0.0(0.0) 0 0 3.0   100%
BH Procedure λ = 0.05 102.2 (9.9) 81 119 4.8 (1.0) 4 6 1.6 (1.4) 0 6 3.2   97%
SAM 0.04 <λ ≤ 0.05 111.5(14.3) 89 129 5.1 (0.6) 5 6 5.6(2.8) 2 12 2.0 6.7 56.5%
  0.03 <λ ≤ 0.04 106.8(13.2) 84 119 3.7 (0.6) 3 5 3.8(2.3) 0 8 1.5 4.0 66.7%
  0.02 <λ ≤ 0.03 96.2(12.5) 80 119 2.3 (0.6) 1 3 3.1(1.7) 1 6 1.4 3.1 39.4%
  0.01 <λ ≤ 0.02 91.0(12.7) 71 107 1.3 (0.47) 1 2 1.6(1.2) 0 4 0.9 1.1 67.5%
  0.00 <λ ≤ 0.01 98.7(6.6) 94 108 0.9 (0.1) 1 1 1.5(1.1) 0 3 1.0 1.9 36.4%
  λ = 0.00 82.9(11.0) 66 108 0.0 (0.0) 0 0 1.0(0.6) 0 3 1.0 1.4 23.1%
RAF 0.04 <λ ≤ 0.05 115.1 (9.2) 96 131 5.1 (0.4) 4 6 4.4(2.7) 1 9 2.2 7.3 75.0%
  0.03 <λ ≤ 0.04 110.6(12.2) 85 128 3.9 (0.6) 3 5 3.2(2.1) 1 8 1.6 3.9 79.2%
  0.02 <λ ≤ 0.03 103.6 (10.6) 86 120 2.7 (0.5) 2 3 2.1(1.5) 0 6 1.3 2.8 81.8%
  0.01 <λ ≤ 0.02 100.7 (10.8) 81 118 1.7 (0.5) 1 2 1.1(0.9) 0 3 0.9 1.3 75.8%
  0.00 <λ ≤ 0.01 100.8 (4.1) 96 112 1.1 (0.2) 1 2 0.7(1.0) 0 3 0.9 1.4 77.8%
  λ = 0.00 83.8 (7.1) 69 95 0.0 (0.0) 0 0 0.1(0.3) 0 1 0.1 0.1 86.2%
  1. FDR, false discovery rate; NGCS, number of genes called significant; ENFP, estimated number of false positives; TNFP, true number of false positives.
  2. | d ¯ | = 1 N λ k = 1 N λ | d k | MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGaciGaaiaabeqaaeqabiWaaaGcbaWaaqWaaeaacuWGKbazgaqeaaGaay5bSlaawIa7aiabg2da9KqbaoaalaaabaGaeGymaedabaGaemOta40aaSbaaeaacqaH7oaBaeqaaaaakmaaqadabaWaaqWaaeaacqWGKbazdaWgaaWcbaGaem4AaSgabeaaaOGaay5bSlaawIa7aaWcbaGaem4AaSMaeyypa0JaeGymaedabaGaemOta40aaSbaaWqaaiabeU7aSbqabaaaniabggHiLdaaaa@445D@ where d k = ENFP K - TNFP K and N λ is number of x <λy in 30 simulations. C ( d 0 ) = k = 1 N λ I k / N λ MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGaciGaaiaabeqaaeqabiWaaaGcbaGaem4qamKaeiikaGIaemizaqMaeyyzImRaeGimaaJaeiykaKIaeyypa0ZaaabmaeaacqqGjbqsdaWgaaWcbaGaem4AaSgabeaakiabc+caViabd6eaonaaBaaaleaacqaH7oaBaeqaaaqaaiabdUgaRjabg2da9iabigdaXaqaaiabd6eaonaaBaaameaacqaH7oaBaeqaaaqdcqGHris5aaaa@428C@ where I k = 1 if d k ≥ 0, otherwise, I k = 0. V a r ( d ) = 1 N λ 1 k = 1 N λ ( d k 0 ) 2 = 1 N λ 1 k = 1 N λ d k 2 MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacPC6xNi=xH8viVGI8Gi=hEeeu0xXdbba9frFj0xb9qqpG0dXdb9aspeI8k8fiI+fsY=rqGqVepae9pg0db9vqaiVgFr0xfr=xfr=xc9adbaqaaeGaciGaaiaabeqaaeqabiWaaaGcbaGaemOvayLaemyyaeMaemOCaiNaeiikaGIaemizaqMaeiykaKIaeyypa0tcfa4aaSaaaeaacqaIXaqmaeaacqWGobGtdaWgaaqaaiabeU7aSbqabaGaeyOeI0IaeGymaedaaOWaaabmaeaacqGGOaakcqWGKbazdaWgaaWcbaGaem4AaSgabeaakiabgkHiTiabicdaWiabcMcaPmaaCaaaleqabaGaeGOmaidaaaqaaiabdUgaRjabg2da9iabigdaXaqaaiabd6eaonaaBaaameaacqaH7oaBaeqaaaqdcqGHris5aOGaeyypa0tcfa4aaSaaaeaacqaIXaqmaeaacqWGobGtdaWgaaqaaiabeU7aSbqabaGaeyOeI0IaeGymaedaaOWaaabmaeaacqWGKbazdaqhaaWcbaGaem4AaSgabaGaeGOmaidaaaqaaiabdUgaRjabg2da9iabigdaXaqaaiabd6eaonaaBaaameaacqaH7oaBaeqaaaqdcqGHris5aaaa@5DBE@ .