ABSTRACT
This chapter reviews applications of quantile regression (QR) to genetic, genomic and other -omic problems. Several aspects make QR an attractive statistical approach in genetic and genomic applications. With the QR approach, instead of sampling from the extremes of the distribution, one can estimate specific parameters related to the extremes with the expectation that it will enhance genetic discoveries. QR is not restricted to the extremes of the distribution and one can estimate regression parameters for any given quantile of the distribution. While most applications of QR concern single genetic marker analyses, recent studies based on multi-marker QR have recently been proposed. In genetic research, QR can provide invaluable insights into associations with specific quantiles of a trait distribution and thus elucidate genetic factors that will not be discovered when looking at the mean of the distribution. Extensions of QR related to high-dimensional data and dependent data would be particularly relevant in genetic and -omic applications.
