A dictionary of genetic effects as a unified representation of the genotype-phenotype map. Eladio J Márquez¹, Rosa Moscarella¹, David Aponte¹, Washington Mio², David Houle¹. 1) Dept of Biological Science, Florida State University, Tallahassee, FL; 2) Dept of Mathematics, Florida State University, Tallahassee, FL.

   We are building a dictionary of genetic effects by systematically manipulating gene expression at target genes, and observing their phenotypic effects. Our model phenotype is the size and shape of the Drosophila wing. We characterize phenotypic responses as the entire multivariate set of responses across the wing, rather than specifying a small number of traits a priori. We manipulate gene expression quantitatively using the mifepristone GeneSwitch system coupled with RNAi-induced transcriptional changes, allowing us to model phenotypic responses as a function of gene expression level. These data probe the genotype-phenotype map, directly providing information about features of the map such as non-linearity of effects and robustness to developmental perturbation, as a function of the role played by a gene in the developmental circuitry. Our approach ensures that the effects inferred for all genes remain comparable, by describing the terms of the dictionary as a difference vector of localized changes relative to a reference wing shape. These vectors provide a common language that facilitates their implementation in any study quantifying the same features, thus providing a powerful tool to link patterns of variation, irrespective of its nature, with putative causal factors whether they are genetic or developmental. We demonstrate practical uses of the Dictionary in a genomic-wide association study, a comparative analysis of interspecies divergence, and an analysis of ecological variation. These examples demonstrate how a catalog of known cause-effect functions can shed light on the direct causation of large-scale phenomena as long as a common, phenomic language is adopted to ensure wide comparability.