Table 2.

Association analysis for C17S4578 comparing case subjects to opposite extreme control subjects in eight linked pedigrees

Strategy	Case subject genotype countsa	Control subject genotype countsa	Dominant model		Recessive model
			Effect size	Benefit-cost ratio	Effect size	Benefit-cost ratio
Strategies focused on age
A. Youngest haplotype-carrier case subject in each pedigree vs. oldest non-haplotype-carrier control subject	5/1/2	1/5/2	1	0.06	11.67	0.73
B. Youngest case subject vs. oldest control subject in each pedigree	4/4/0	1/3/4	16.0b	1.0	7	0.44
Strategies focused on low-covariate status
C. All low-covariate haplotype-carrier case subjects vs. all high-covariate non-haplotype-carrier control subjects	7/5/1	0/4/6	18c	0.78	23.3b,d	1.01
D. Families with 2+ low-covariate haplotype-carrier case subjects vs. all high-covariate non-haplotype-carrier control subjects	4/4/1	0/4/6	12c	0.63	16.0b,d	0.84
E. All low-covariate case subjects vs. all high-covariate control subjects	7/6/3	0/4/7	7.58	0.28	17.1b,d	0.63
F. Lowest covariate case subject vs. highest covariate control subject per pedigree	3/4/1	0/4/4	7	0.44	5.3	0.33
Other strategies
G. All haplotype-carrier case subjects vs. all non-haplotype-carrier control subjects per pedigree	10/21/9	7/66/104	4.91c	0.02	8.1c	0.04
H. All case subjects vs. all control subjects	10/40/31	10/86/110	1.85c	0.01	2.76c	0.01
I. Random case subject vs. random control subject per pedigree	0/5/3	1/2/5	2.78	0.17	0	0.00

^a The three values are the counts for homozygote rare, heterozygote, and wild-type variants.

^b For cells with zero count, we added 0.5 to the cell to compute the odds ratio.

^cp < 0.05.

^d Unable to compute significance because of a zero cell, and either case or control subjects or both are related.