TABLE 3.

Correspondence Between Key Results From Easton et al. [2007] and This Analysis

From Easton et al. [2007], Table 5			Reparsing based on this work
Missense or in-frame indel, by domain	n	α (95% CI)	Missense substitutions, by domain	Class	n	Joint risk estimatea	α (95% CI)
BRCA1 BRCTor BRCA2 DBD			BRCA1 RINGb or BRCTor BRCA2 DBD	C65	98	3.12	0.81 (0.61-0.95)
				C55	15	2.28	0.66 (0.34-0.93)
				C45	12	2.14
	323	0.35 (0.26-0.45)c		C35	28	1.93
				C25	28	1.67	0.29 (0.09-0.56)
				C15	30	1.36
				C0	242	1.16	0.01 (0.00-0.06)d
Not BRCT nor DBD	854	0.00 (0.00-0.04)c	Not BRCA1 RING, BRCT	GD = 0	554	1.04	0.01 (0.00-0.04)d per Easton etal. [2007]
			Not BRCA2 DBDe	GD > 0f	677	1.05

^aCalculated by Gaussian smoothing.

^bExcluding substitutions at M1 and the eight canonical C3HC4 residues.

^cPlease note that the heterogeneity analysis calculations inTable 5 of Easton et al. [2007] included some in-frame insertion deletion variants and also included data from segregation analyses. Recalculation for RING, BRCT, and DBD missense substitutions (and excluding substitutions at M1 and the eight canonical C3HC4 residues) using the FamHx-LR alone yielded a heterogeneity point estimate of 0.32. The point estimate for the set of missense substitutions lying outside of the RING, BRCT, and DBD domains remained 0.00.

^dIn fact, the FamHx-LR heterogeneity analysis point estimate for this Class is 0.00. However, these point estimates will serve as prior probabilities in future analyses of unclassified BRCA1 and BRCA2 missense substitutions. Because it would be impossible to modify a prior probability of 0.00, we report these point estimates as 0.01.

^eAnd also excluding substitutions at BRCA2 M1 and substitutions that fall within 2 bp of a splice junction.

^fA detailed study of the potential genetic risk due to missense substitutions in the remaining small well-conserved BRCA1 and BRCA2 sequence elements will be the subject of a future study.