Odds of observing an iKIR modification across <i>DQA1*01</i>:<i>02</i>, <i>DQB1*03</i>:<i>01</i>, <i>DQA1*02</i>:<i>01</i> and <i>DQA1*01</i>:<i>02-DQB1*06</

<p>The iKIR effect observed in the case-control cohort was validated in an independent family dataset. Trios were stratified into high (<b>></b>threshold) and low (<b>≤</b> threshold) iKIR score according to the iKIR score of the child in each trio. For each threshold and each genotype, we calculated the ratio of transmitted to non-transmitted genes in each stratum. The odds of observing an equal or greater difference between log ratios across 4 frequent protective genotypes (<i>DQA1*01</i>:<i>02</i>, <i>DQB1*03</i>:<i>01</i>, <i>DQA1*02</i>:<i>01</i> and <i>DQA1*01</i>:<i>02-DQB1*06</i>:<i>02</i>) were assessed by permutation test.</p> <p>(PDF)</p>