Cells are not just bags of molecules—they are dynamic networks where proteins switch on and off, and no current technique can reliably say which proteins are active when or which interact with which inside a living cell. This matters because sequencing genomes has become routine, but knowing a cell’s parts list is not the same as understanding how those parts work together. Without predictive models of cellular networks, drug development, biomanufacturing, and even basic biology remain hit-or-miss. The researchers argue that the field needs to start small: focus on tractable systems like the bacterial chemosensory pathway—a well-understood signalling circuit—build a reliable model, test it, and then extend it to more complex pathways in higher organisms. If this approach succeeds, it could transform how we design antimicrobials. By identifying signature motifs in signalling networks, researchers might find new drug targets that disrupt bacterial behaviour without killing the cell outright—potentially slowing resistance. The work is fundamental science, not a product. But similar bottom-up modelling in physics and engineering has repeatedly led to unexpected breakthroughs. A predictive model of a cell’s dynamic wiring would be a tool, not a treatment—one that could eventually accelerate drug discovery, improve fermentation in food production, and guide bioremediation.
View original technical description
Life arises not just from individual molecules, but from their dynamic interactions. A great deal of time and money has been spent globally on sequencing genomes, followed by high-throughput methods for transcriptomic, proteomic and metabolomic measurement. Cells, however, are complex systems of networks and pathways. A long term goal of molecular biology must be to take a sequenced genome and be able to model the intracellular activities of a cell, and ultimately a complex organism. The availability of accurate and robust models has the potential to enable major advances in a range of activities ranging through pharmaceuticals and health care to food production and bioremediation. Approaches can be top down, bottom up or through accurate definition of limited inputs and outputs whilst ignoring the intervening detail. The goal is an accurate model of the dynamic activities of a cell under a specific set of growth conditions, when many proteins may be present but not active, only interacting with target proteins or DNA when activated to do so (the so-called 'sensome' or 'interactome'). None of the techniques available to date can reliably say which proteins are active when or which interact with which. Even recent developments in, for example, fast through-put guided yeast two hybrid systems tell you what can interact, not what does interact in vivo and when. We believe the best way forward is to initially focus on a limited number of tractable and important biological systems, for example, the bacterial histidine protein kinase (HPK) dependent chemosensory pathway or a defined section of the yeast cell cycle pathway, and to develop a reliable, predictive model for those systems. With, for example the bacterial chemosensory pathway, models can be tested through to species with complex pathways and these then extended to the total complement of HPK pathways in selected bacterial species. The long term goal is to carry these approaches to more complex pathways in higher organisms, and eventually to extend the models to sensory systems in general, with each model tested biologically as the programme develops. These studies will lead not only to an understanding of the complex interactive networks of signalling pathways, but also to the identification of signature motifs, possibly providing new targets for antimicrobials.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know