UCalgary Biostatistics Seminar: Jinko Graham

Linkage analysis maps genes for a heritable trait by identifying genomic regions with excess relatedness among individuals with similar trait values.Analysis may be conducted in families, or in unrelated individuals from a population. Here, we focus on linkage analysis in a population sample, but use sequences rather than individuals as our unit of observation. We propose two ways to associate similarity in relatedness of sequences with similarity
in their trait values and compare the resulting linkage methods to two genotypic-association methods. We also introduce a procedure to label case sequences as potential carriers of causal variants, based on inferred relatedness to other case sequences. Our simulation results indicate that linkage methods based on sequence-relatedness improve gene localization and perform at least as well as genotypic-association methods for detecting rare causal variants within a gene for an allelically heterogeneous disease trait.