Table 3. Genotype counts and their transmissions used by .

		Nontransmitted							Nontransmitted
		haplotype							haplotype
Transmitted							Transmitted
haplotype		AB	Ab	aB	ab	Total	haplotype		AB	Ab	aB	ab	Total
	AB		30	3	7	40		AB		24	5	5	34
	Ab	37		6	5	48		Ab	31				31
	aB	8	6		3	17		aB	9				9
	ab	11	3	1		15		ab	11				11
	Total	56	39	10	15	121		Total	51	24	5	5	85

Haplotypes in rows represent those transmitted haplotypes at each genotype. Haplotypes in columns represent those nontransmitted haplotypes at each genotype. Homozygous genotype counts (diagonal) are crossed off the tables as they are not used to compute Left grid: genotype counts from the training data set (see Table 2) used to make up groups and in Groups are: with those haplotypes with counts larger than counts (Ab: versus and aB: versus ) and with counts larger than counts ( versus ). Right grid: genotype counts from the test data set used to compute the statistic. As the length similarity measure is used to assign an haplotype to a group, and the two most similar haplotypes to haplotype belongs to group is assigned to All the haplotypes belonging to the same group are considered of having an equivalent effect and are collapsed. Therefore, parental genotypes in the test data set with haplotypes belonging to the same group are considered as homozygous and not used by (they are crossed off the table too).