Longitudinal linear combination test for gene set analysis

BMC Bioinformatics

Table 2 The number of significant gene sets found by LLCT at different levels of confidence, testing a variety of outcomes and datasets

Datasets	Type I Error	SBP	DBP	SBP& DBP^a	SBP-DBP^b	HTN
Adjusted for smoking status
Related Subjects	1%	30	23	20	73	65
	5%	170	135	141	360	321
	10%	255	278	310	434	392
Unrelated Subjects	1%	12	3	5	27	5
	5%	136	39	60	389	82
	10%	408	78	245	735	162
Adjusted for antihypertensive medications
Related Subjects	1%	98	13	63	127	12
	5%	402	127	271	541	99
	10%	413	242	390	614	159
Unrelated Subjects	1%	17	3	11	17	2
	5%	142	60	86	116	22
	10%	465	75	186	382	88
No Adjustment
Related Subjects	1%	18	17	14	43	54
	5%	158	141	122	259	327
	10%	263	273	277	386	417
Unrelated Subjects	1%	9	2	3	17	2
	5%	234	37	70	273	71
	10%	537	68	231	682	168

^aThe multiple analysis of systolic and diastolic blood pressure. In this analysis, the outcome is a linear combination of SBP and DBP with the highest association with the linear combinations of gene expressions
^bPulse pressure which is the difference between systolic and diastolic blood pressures

ISSN: 1471-2105