Statistical methods for testing X chromosome variant associations: application to sex-specific characteristics of bipolar disorder

Table 1 Different coding schemes for the SNP variable reflect different assumptions regarding XCI status

Assumed XCI status	Coding scheme	Coding of the SNP variable (Eq. 1)				Calculation of ORs^a
		Sex	# copies of effect allele			Calculation of ORs^a
			0	1	2	OR_M	OR_W1	OR_W2
Subject	Clayton	M	0	2	NA	\( {e}^{2{\beta}_1} \)	\( {e}^{\beta_1} \)	\( {e}^{2{\beta}_1} \)
Subject	Clayton	F	0	1	2	\( {e}^{2{\beta}_1} \)	\( {e}^{\beta_1} \)	\( {e}^{2{\beta}_1} \)
Escape	PLINK	M	0	1	NA	\( {e}^{\beta_1} \)	\( {e}^{\beta_1} \)	\( {e}^{2{\beta}_1} \)
Escape	PLINK	F	0	1	2	\( {e}^{\beta_1} \)	\( {e}^{\beta_1} \)	\( {e}^{2{\beta}_1} \)

^aExpressions for calculating the OR are shown for the effect of one allele copy in men (OR_M) and one or two copies in women (OR_W1 and OR_W2). For men, but not women, the OR is calculated differently based on whether the SNP is believed to be subject to XCI (and hence the coding scheme, PLINK or Clayton, that is chosen)

ISSN: 2042-6410