Application Studies Of Several TDT Expansion Methods For Family-based Association Study With Missing Single Parental Genotype

Transmission disequilibrium test(TDT)is an association analysis method based on linkage disequilibrium,and it is the gold standard for testing the association between genetic variation and disease against core samples composed of infected individuals and their parents.Note that complex pedigree structures,that is,siblings without parents,or three-generation pedigree structures that may be missing genotypes,are the most common in reality.There are several extension methods of TDT that can be applied to any pedigree structure.In this paper,we consider three common extension methods:(1)Pedigree disequilibrium test(PDT): PDT can take advantage of all the information of extended pedigrees without having to divide them into multiple core families,nor do they need all genotypes to be available.(2)Family-based association test(FBAT): FBAT is a method to calculate the expected value of various association test statistics under the null hypothesis of linkage but no association,and it can use the genetic model of markers within and between families to test the significance of the associations.(3)Generalized disequilibrium test(GDT): GDT uses all uncoordinated relative genotype differences to assess the association test within the family.The three methods all have corresponding software,which can be used to compare their effectiveness and statistical power under three different environment settings.Environment 1(U): all founder genotypes are independently distributed,meet the Hard-Weinberg equilibrium(HWE)hypothesis,and have the same minor allele frequency(MAF).Environment 2(HWE-D): All founder genotype distributions have the same MAF,but genotype distribution of the founder deviates from the expected HWE ratio,and the inbreeding coefficient is F(28)03.0.Environment 3(ADM): Simulates the effects of population mix.We used simulated data to study the performance of three TDT expansion methodsin the presence of population exogenous,genotype distributions that are not in HWE,and different pedigree structures and linkages.Our results show that the type I error rate of some TDT extension methods will be significantly higher than the nominal level.FBAT and GDT are more and PDT is not suitable for pedigree with deletion of parental genotypes.In order to analyze the rare variation in the next generation sequencing,the region-based TDT extension method is comprehensively considered,and it is found that the extension method depends on the robust assumption of unit point test.Finally,we compare and analyze several mixed model methods.