Modelling

Structural modelling can be used to generate hypotheses about the structure (and therefore to imply things about the biochemical function) of macromolecules. The recent release (2021) of the AlphaFold algorithm by Google DeepMind has been a major step forward in proteins 3D structure modelling starting from their sequences: AI researcher Demis Hassabis and chemist John M. Jumper were awarded the 2024 Nobel Prize in Chemistry for AlphaFold protein prediction, sharing it with Prof. David Baker (recognised for his contribution to computational protein design.) In a joint effort between EMBL-EBI and Google DeepMind, AlphaFold predictions have been made publicly available on the AlphaFold Protein Structure Database (AlphaFold DB). If you are interested in diving deeper into AlphaFold, try our online tutorial or watch this recorded webinar. 

Moving from the single macromolecules to their interactions, biologists are often interested in process modelling (a branch of systems biology). A review by Bartocci and Lió provides a good summary of the kinds of processes that are amenable to systems modelling, and some examples of tools that enable you to do this. You can download mathematical models in a number of standard formats from the BioModels Database. Check out our Biomodels: Quick tour to learn more about this.

Systems modelling forms part of a cycle in which experimental biology plays an equally important role: a typical systems modelling cycle involves building a model that represents what you know about the biology then testing to see whether it behaves in the same way as the biological system itself. If not, you tweak the model to come closer to the biological system and repeat the cycle until your model faithfully represents reality.

Try our guided example on structural modelling of IRAK2.