Table 1.

Allele counts (frequencies) of the risk/ non-risk alleles [n (%)] of the CNV near UXS1

absolute chromosomal position (hg18)	relative position to nearest gene(s)	Part of study	n (%) –risk allele in PsA	n (%) – non-risk allele in PsA	n (%) –risk allele in Ctrl	n (%) – non-risk allele in Ctrl	P-value	OR [95% CI]
chr2: 106,246,527–106,251,789	69 kb upstream of UXS1	discoverya	861 (90.1)	95 (9.9)*	6026 (79.3)	1570 (20.7)*	2.84 × 10−15	2.36 [1.90–2.94]
		unfiltered discoveryb	635 (66.4)	321 (33.6)	5145 (67.7)	2451 (32.3)	0.44	0.94 [0.82–1.09]
		replication	331 (66.5)	167 (33.5)*	594 (66.0)	306 (34.0)*	0.94	1.02 [0.81–1.29]

Deletion is marked by *. The p-value corresponds to a χ2 test of 478 PsA patients and 3798 control individuals. Odds ratios [95% confidence interval] of the discovery study (array-based analysis) were calculated for the same cohorts. The independent replication cohorts (MLPA-based analysis) comprised 251 patients and 451 control individuals. ^aDiscovery describes the initial SNP array-based analysis with the filter criteria of 5 markers and 5 kb before validation, ^bunfiltered discovery the number of CNVs and wildtype-alleles in the initial SNP array-based dataset that was unfiltered for no. of markers and for size, as analyzed after validation with MLPA