During the patterning of the vertebrate anterior-posterior axis, a gene expression oscillator controls segmentation of the somites, the precursors of the vertebral column. Several mathematical models have been developed to explain how this segmentation clock works. Now, Ahmet Ay, Ertuğrul Özbudak and colleagues () extend these earlier models by developing a multicellular stochastic computational model of the zebrafish segmentation clock. Simulations run with this model show that autoregulatory negative-feedback loops of dimers of Hairy/enhancer-of-split-related (Hes/Her) proteins - transcriptional repressors that are encoded by the only conserved oscillating genes in vertebrates - can drive synchronised gene expression oscillations in wild-type zebrafish embryos and can recapitulate various mutant phenotypes. The model also predicts that synchronised oscillations can only be generated if zebrafish Her proteins have a half-life of less than 6 minutes, a prediction that the researchers validate by showing that zebrafish Her7 has a half-life of 3.5 minutes. Importantly, this model can now be used to make predictions about clock behaviour that can be tested experimentally.
