A computational science case study: classification of hybrids using genetic markers and maximum-likelihood estimates

A self-contained undergraduate level "case study" in computational science/biology is described. The application presented brings together genetics, statistics, and numerical methods into a unified course project. The project we describe involves a statistical model and genetic classification of individuals that may be the result of hybridization between genetically divergent parents. Actual data consist of inherited genetic markers, which allow the evaluation of alternative model parameters by a maximum-likelihood technique. In the implementation of the classification model, the project provides opportunities for the use of numeric computing in C/C++ using libraries like the GNU Scientific Library for the computation of the model parameters.