Skip to main content

Table 5 Modification of the results obtained in the GRIV case-control study when using the various subhaplotyping methods

From: Computation of haplotypes on SNPs subsets: advantage of the "global method"

Genes

Sub-haplotype

p-value direct Rmax subhap

p-value global Rmax subhap

p-value Combination Rmax subhap

IL10Receptor

Exon

0.026

A.H cases: 100%

A.H controls: 100%

*0.103

A.H cases: 100%

A.H controls: 100%

0.093

A.H cases: 88%

A.H controls: 99%

IL4Receptor

Promoter

0.019

A.H cases: 100%

A.H controls: 100%

*0.072

A.H cases: 100%

A.H controls: 100%

*0.088

A.H cases: 100%

A.H controls: 98%

IL6

Promoter

0.059

A.H cases: 100%

A.H controls: 100%

0.012

A.H cases: 100%

A.H controls: 100%

*0.009

A.H cases: 82%

A.H controls: 90%

  1. *Best method regarding the missing data level.
  2. This table presents the p-values found for the Fisher's exact tests comparing the subhaplotypes distributions between seropositive patients of the GRIV cohort (cases) and seronegative subjects (controls). The subhaplotypes were computed either with the direct Rmax method as previously published [2, 43], or with the global Rmax method, or with the combination Rmax method described in our study. The percentage of missing information was respectively 6.7%, 11.1% and 14.8% for the IL10R, IL4R, and IL6 genotypic data. One can see that some signals which were previously published as positive (p < 0.05) using the direct method become negative, while some signals which were previously published as negative become much stronger thanks to the novel subhaplotyping methods. For IL10R we have a deficit of information for cases and as consequences a lower percentage of assigned haplotypes in the combination method which is more restrictive.
  3. A.H cases: percentage of assigned haplotypes attributed in the tests for cases
  4. A.H control: percentage of assigned haplotypes attributed in the tests for controls.