Experimental Determination of Drosophila Embryonic Coordinates by Genetic Algorithms, the Simplex method, and Their Hybrid

Modern large-scale \functional genomics" projects are inconceivable without the automated processing and computer-aided analysis of images. The project we are engaged in is aimed at the construction of heuristic models of segment determination in the fruit fly Drosophila melanogaster. The current emphasis in our work is the automated transformation of gene expression data in confocally scanned images into an electronic database of expression. We have developed and tested programs which use genetic algorithms for the elastic deformation of such images. In addition, genetic algorithms and the simplex method, both separately and in concert, were used for experimental determination of Drosophila embryonic curvilinear coordinates. Comparative tests demonstrate that the hybrid approach performs best. The intrinsic curvilinear coordinates of the embryo found by our optimization procedures appear to be well approximated by lines of isoconcentration of a known morphogen, Bicoid.