Statistical analysis with missing data
Statistical analysis with missing data
Application of Dempster-Schafer Method in Family-Based Association Studies
IEEE/ACM Transactions on Computational Biology and Bioinformatics (TCBB)
| Hi-index | 0.03 |
The Transmission Disequilibrium Test (TDT) detects linkage between a marker and a disease-susceptibility locus in the presence of linkage disequilibrium. The TDT requires data on the genotypes of affected offspring and their parents, which might not always be available. For example, for late onset diseases it might be difficult to find parents still alive, or genotypes of offspring might not be available. Genotyping unaffected siblings, combining different genotype data sets, or assuming a model mechanism for missing parents have all been proposed to deal with missing genotypes in parents but not in offspring. In this paper, we propose a Mendel Inheritance-Transmission Disequilibrium Test (MI-TDT) to impute missing genotypes in any members of a family with two affected offspring. Our method does not require any of the remedies mentioned above but simply utilizes the fundamental property of Mendel Inheritance on the transmission of alleles from parents to offspring. Most importantly, the MI-TDT reassures researchers about the declared significant genes when incomplete data is ignored. We illustrate the MI-TDT by identifying significant genes in type 1 diabetes from the Warren families in the United Kingdom.