We study gene expression data, derived from developing tissues, under multiple genetic backgrounds (mutations). Motivated by the perceived behavior under these background, our main goals are to explore time windows questions: 1. Find a large set of genes that have a similar behavior in two different genetic backgrounds, under an appropriate time shift. 2. Find a model that approximates the dynamics of a gene network in developing tissues at different continuous time windows. We first explain the biological significance of these problems, and then explore their computational complexity, which ranges from polynomial to NP-hard. We developed algorithms and heuristics for the different problems, and ran those on synthetic and biological data, with very encouraging results.