Bayesian metamodeling of complex biological systems across varying representations
© mattjeacock/DigitalVision Vectors/Getty Images
A more precise and complete understanding of cell structure and function could result from a new way of combining different types of biological models into a single mathematical framework.
Most cell-modelling approaches focus on specific biological processes. Hence, each model can capture just a small part of the complex inner workings of the cell.
Now, researchers from ShanghaiTech University and elsewhere have developed a system for integrating these different types of models. They produced a ‘metamodel’ that can create a more accurate and comprehensive cellular map by assembling and harmonizing disparate input models.
To demonstrate the power of this approach, the researchers built a metamodel of how beta cells in the pancreas secrete insulin in response to sugar cues — a process that goes awry in people with diabetes.
The method could offer new insights into biological mechanisms to inform drug-discovery efforts.
- PNAS 118, e2104559118 (2021). doi: 10.1073/pnas.2104559118