Table 2. Logistic regression analysis of identified CNVRs with risk of type 2 diabetes in Korean subjects (n = 771).
| CNV region* | Length (kb) | Cytoband | Genes (Nearby genes) | Type | Case (Freq.**) | Control (Freq.**) | OR (95% CI) | P | Pcorr † | DGV |
| chr15:45994758–45999227 | 4.5 | 15q21.1 | (SEMA6D, SLC24A5) | Deletion | 39(7.86%) | 39(14.18%) | 2.6 (1.5–4.5) | 8.6E-04 | 0.01 | Overlapped |
| chr22:20722473–21702142 | 979.7 | 22q11.22 | GGTLC2,LOC648691,LOC96610,POM121L1P,PRAME,VPREB1,ZNF280A,ZNF280B | Deletion | 27(5.44%) | 4(1.45%) | 0.2 (0.7–0.1) | 0.009 | 0.06 | Overlapped |
| chr18:3559620–3561217 | 1.6 | 18p11.31 | DLGAP1 | Deletion | 45(9.07%) | 43(15.64%) | 2.0 (1.2–3.4) | 0.01 | 0.06 | Overlapped |
Abbreviations: CNV (copy number variation), OR (odds ratios), CI (confidential interval), Pcorr (corrected P value), DGV (Database of Genomic Variants).
*The version of human reference genome: NCBI build 36/hg18.
**The frequency was obtained by number of identified CNVs/number of case subjects in case group and number of identified CNVs/number of control subjects in control group.
†
The false discovery rate (FDR) method used to perform a multiple testing of copy number variation regions (CNVRs) (Freq.>1%).
Bold values indicate the case of P<0.05.