Skip to main content

Table 3 Case-only χ 2 test for interaction between KIR3DS1 - KIR3DL1 and HLA-Bw4, across the ten multiply imputed qPCR and SNP datasets

From: A hybrid qPCR/SNP array approach allows cost efficient assessment of KIR gene copy numbers in large samples

a)

qPCR

SNP

  

HLA-Bw4-

HLA-Bw4+

HLA-Bw4-

HLA-Bw4+

KIR3DS1 -

KIR3DL1 +

183

269

739

1063

KIR3DS1 +

KIR3DL1 -

12

21

40

71

KIR3DS1 +

KIR3DL1 +

113

138

396

613

   

p-value =0.4094

 

p-value =0.4235

b)

qPCR

SNP

  

HLA-Bw4 -

HLA-Bw4 +

HLA-Bw4 -

HLA-Bw4 +

KIR3DL1 -

12

21

40

71

KIR3DL1 +

296

407

1135

1676

   

p-value =0.5144

 

p-value =0.3609

c)

qPCR

SNP

  

HLA-Bw4-80I -

HLA-Bw4-80I +

HLA-Bw4-80I -

HLA-Bw4-80I +

KIR3DS1 -

293

159

1153

649

KIR3DS1 +

159

107

673

447

   

p-value =0.4922

 

p-value =0.0353

  1. Counts in each contingency table are derived from the most likely copy number assignment across the multiply imputed datasets. To reduce the degrees of freedom and improve power, we summarise copy numbers higher or equal to one by presence (+) and zero by absence (-). The pooled p-value of each χ2 test, across the multiply imputed datasets, is given in the last row of each contingency table. We find no significant association with HLA-Bw4, within cases, in either the joint (a) or the marginal (b) (c) KIR3DS1-KIR3DL1 distributions.