| dc.description.abstracteng | The Mendelian genetics concept relies on simple genotype-phenotype relationships with few genes having major effects on the phenotype. However, many traits evolve by a combination of variation in many genomic loci with minor phenotypic effects. Therefore, it is challenging to reveal individual genes underlying natural variation in quantitative traits. Most genes do not act individually, but they are interconnected in gene regulatory networks (GRNs). Revealing variable nodes and modules within GRN, thus has the potential to gain mechanistic insights into phenotypic evolution. The insect head that harbours the compound eyes is a complex quantitative trait that is highly variable in Drosophila. The formation of the insect compound eye is determined by a complex GRN composed of more than 5,000 genes. To reveal the molecular and developmental basis of natural variation in eye size and head shape, I studied head development in Drosophila melanogaster and D. mauritiana. Eye size varies in these two species due to differences in ommatidia number and a trade-off between eye size and interstitial head cuticle has been observed. To reveal novel candidate genes, I integrated several unbiased genome wide datasets, such as developmental gene expression (RNAseq), chromatin accessibility (ATACseq) and quantitative trait loci mapping data. This integrative approach unravelled 65 candidate genes, which I validated functionally for their functional involvement in eye development applying an RNA interference screen. Phenotypically relevant candidate genes were used to reconstruct a novel GRN module that contains predominantly genes with variable expression between species. The addition of few extra genes to this network allowed me to propose developmental processes that may be variable. I tested one of these hypotheses functionally to show that Jim, Pnr and Upd are co-expressed during head and eye development, suggesting a novel role of Jim during this process. Overall, my finding shows that a GRN-centric approach is highly powerful to reveal the mechanisms underlying the evolution of complex organ development. | de |