BMC Medical Research Methodology

Table 3 Operating characteristics of four methods under eight scenarios

From: IBIS: identify biomarker-based subgroups with a Bayesian enrichment design for targeted combination therapy

	Simulation 1		Simulation 2
Method	FWER	Conjunctive Power	FWER	Conjunctive Power
	Scenario 1		Scenario 1
IBIS	0.0971	/	0.0490	/
Independent	0.1000		0.0496
BHM	0.0991		0.0500
Freq	0.0985		0.0499
	Scenario 2		Scenario 2
IBIS	/	0.8602	/	0.8092
Independent		0.6089		0.4664
BHM		1.0000		1.0000
Freq		0.8961		0.8787
	Scenario 3		Scenario 3
IBIS	0.0455	0.8984	0.0265	0.8193
Independent	0.0916	0.6447	0.0449	0.4910
BHM	0.3093	0.6027	0.1760	0.4587
Freq	0.2149	0.8609	0.1717	0.7582
	Scenario 4		Scenario 4
IBIS	0.0578	0.9541	0.0399	0.9341
Independent	0.0079	0.3748	0.0045	0.2251
BHM	0.7718	0.9980	0.6805	0.9964
Freq	0.1072	0.8101	0.0866	0.7900
	Scenario 5		Scenario 5
IBIS	0.1586	0.8774	0.1106	0.8393
Independent	0.0693	0.6200	0.0343	0.4743
BHM	0.6445	0.9768	0.4459	0.9407
Freq	0.1772	0.8480	0.1647	0.8375
	Scenario 6		Scenario 6
IBIS	0.1178	0.8092	0.0838	0.7779
Independent	0.0349	0.3766	0.0175	0.2201
BHM	0.7028	0.9891	0.5440	0.9686
Freq	0.1153	0.7656	0.0973	0.7508
	Scenario 7		Scenario 7
IBIS	0.1877	0.8108	0.1508	0.7742
Independent	0.0525	0.2507	0.0248	0.1185
BHM	0.6700	0.9688	0.4945	0.9184
Freq	0.1860	0.7138	0.1712	0.7012
	Scenario 8		Scenario 8
IBIS	0.0843	0.8617	0.0386	0.7637
Independent	0.0079	0.4113	0.0045	0.2171
BHM	0.4294	0.9982	0.3169	0.9937
Freq	0.3961	0.9030	0.3030	0.8597

Simulations 1 & 2 correspond to different degrees of control over FWER in the global null scenario (scenario 1)
For global null scenario (scenario 1), calculating the conjunctive power makes no sense. Similarly, for global alternative scenario (scenario 2), calculating the FWER makes no sense

Back to article page

ISSN: 1471-2288

Contact us

General enquiries: journalsubmissions@springernature.com