Unleashing the potential of whole-cell modelling one step at a time

Whole-cell models (WCM) that mechanistically represent every gene and gene function have many potential applications, from knowledge discovery to addressing real-world problems in engineering and medicine. However, there are many challenges to building, simulating and validating WCMs. It would be more realistic to achieve each cell-specific WCM as a long-term goal, where each model is built in stages and released as incremental versions. This strategy is exemplified by recent WCMs for two bacterial species, each of which has been continuously expanded since the release of their first model versions. Each expansion improved the models but also revealed biological insights about cells. This demonstrated the values of WCMs and the benefits (to both authors and the wider community) of disseminating organism-specific models as incremental versions. In this review, I compare existing bacterial WCMs and extract a set of minimal features that would be attainable in the short term while maintaining the distinctive abilities of WCMs. This minimal set could be used as guidance for initial versions of future WCMs, such as for organisms with large genomes. I discuss different collaborative formats and what we would need for community-driven development of WCMs. I also discuss how we can leverage emerging technologies in artificial intelligence and computing to accelerate whole-cell modelling.