Search
next up previous
Next: Simon Tavaré - The Up: Mathematical Genetics and Genomics Previous: Nicholas Schork - The

Katy L. Simonsen - Probability models for genetic factors underlying a binary phenotype



KATY L. SIMONSEN, Statistics Department, Purdue University, West Lafayette, Indiana  47905-1399, USA
Probability models for genetic factors underlying a binary phenotype


Probability models in quantitative genetics have a long history. Models for linkage between a single binary trait locus ( BTL) and a single genetic marker, and for two or three BTL and marker loci have been developed for specific experimental designs such as backcross and F2. However, these models do not generalize easily to other designs, and do not allow for varying penetrance with multiple BTL. The development of a more general model with explicit parametrization for incomplete penetrance and multiple genes would permit the expansion of detection methodology into other experimental designs. In this work a general probability model for linkage between an arbitrary number of BTL and marker loci is developed. This model incorporates varying penetrance, and allows for any experimental design as long as the gametic frequencies of the originating parental lines are known. The generality of the formulation streamlines the computer implementation of statistical methods for locating BTL. The model easily reduces to well known special cases, and the joint probability distribution can be used to generate expected values and conditional probabilities for any experimental population derived from the initial cross of two different inbred parental lines. This model allows further practical developments for estimation and testing, including BTL detection, and unbiased estimation of both penetrance and BTL location.


next up previous
Next: Simon Tavaré - The Up: Mathematical Genetics and Genomics Previous: Nicholas Schork - The