Local adaptation, ecological divergence, and speciation are now accepted to occur over short timescales, yet we still have limited understanding of why some lineages adapt rapidly while others do not. Growing evidence suggests that genome architecture, such as physical variation, structure, and repetition, and its effects on the expression of coding and noncoding gene products are primary determinants of evolutionary potential. Insect-plant interactions are an ideal system for investigating these mechanisms. Insects comprise ~55% of described eukaryotic species, roughly half of which are herbivorous, making plant-feeding insects nearly a quarter of all known eukaryotic life. Most insects are specialists restricted to few plant families. But host repertoires are not static — they shift across ecological and evolutionary timescales. We seek to uncover the genomic and transcriptomic architecture underlying these host-use patterns. Critically, these mechanisms must be situated within the multitrophic context that defines an organism's niche, including host plant chemistry, nutrition, and associated microbial communities, as these ecological pressures ultimately shape host plasticity, host shifts, and biodiversity. We investigate these questions in several insect systems, focusing on Tephritis conura, its coevolved bacterial endosymbiont, and Cirsium thistle hosts. We use genomic, transcriptomic and metabolomic data to investigate genome evolution in Tephritid flies, ecotype divergence and convergent host adoption, and multitrophic interactions underlying specialization and multiple host use. We also investigate the genomics and transcriptomics of multiple host use in several butterfly systems, including Pieris spp. responding to invasive mustards and herbicides and Nymphalids with different host repertoires.