Plant & Animal Genomes XI Conference Plant & Animal Genomes XI ConferenceJanuary 11-15, 2003
Town & Country Convention Center
San Diego, CA
Bioinformatics: Software
Computer: Poster and Demo
Reduction of a pairwise-genetic-linkage matrix to a chromosome map requires means of proposing candidate orders and fitting multilocus distances. We may do both by treating genetic loci as atoms interacting in low-dimensional space by classical mechanics. A linear force function of the error between raw and fitted interlocus map distance, weighted by linkage information, leads to an energy statistic identical to a least-squares test of fit. Iterated dynamics calculations lead to 1-, 2-, and 3D configurations interpretable as genetic maps. Linear distance estimates may then be updated with locus-order information by the Lander–Green or other algorithms. QMap is a Java application that uses this physical analogy to construct maps with raw genotype data from common plant mating designs. Its graphical displays track convergence and respond dynamically to user-imposed parameter changes. The distribution of force and energy across finished maps can give information about crossover interference and the spatial structure of chromosomes, and about the influence of data quality on map quality.