Cellular networks interact to result in the organismal phenotype, yet these networks have so far been difficult to integrate. Chandrasekaran and Price have developed an approach — Gene Expression and Metabolism Integrated for Network Interference (GEMINI) — that is able to constrain predicted gene regulatory networks on the basis of metabolic data in yeast. They used GEMINI to build a network in Saccharomyces cerevisiae and used it to predict phenotypes such as growth effects after transcription factor knockout in new conditions. This will be a valuable tool in synthetic biology.