My research is driven by an interest in the complex relationship between genotype and phenotype. To gain a mechanistic and quantitative understanding of the genotype-phenotype map is one of the big challenges in biology today. Tackling this challenge requires a quantitative systems-level understanding of the gene networks underlying development across multiple levels, from the molecular to the organismic. This is difficult due to the large number of factors involved. We depend on computational models for this task. I present a reverse-engineering approach, where gene regulatory interactions are inferred from quantitative expression data, using data-driven mathematical models (called gene circuits). We have established that the gap gene network can be consistently reconstructed in this way using both protein or mRNA expression data. Gap gene circuit models in Drosophila reproduce observed gene expression with high precision and temporal resolution and reveal a dynamic mechanism for the control of positional information through shifts of gap gene expression domains. My group is extending this approach to a comparative study of the gap gene network between different species of dipterans. No such quantitative systems-level analysis of an evolving developmental gene regulatory network has been achieved to date. We have created and analyzed quantitative data sets for gap gene expression in the scuttle fly Megaselia abdita, and the moth midge Clogmia albipunctata, which are now used for model fitting. Our approach yields precise, quantitative predictions of how changes of gene regulatory feedback affect the timing and positioning of expression domains in these species. These predictions are now being tested experimentally using RNA interference.