https://www.sciencedirect.com/science/article/pii/S2352914820301714
A team of members from the University of the West of England and the University of Novi Sad have come out with a paper in Informatics in Medicine Unlocked that talks about a new approach to optimize a simulated targeted drug delivery system.
By utilizing hybrid fitness equations that run based on how different (novel) a solution they can come up with, eventually it will lead to an evolutionary selection of the best strategy. This allows them to outperform conventional objective-based (greedy) searches for solutions, which drive towards a single outcome despite potential dead-ends in logic.
Applying this novelty search to PhysiCell, a tumour simulation program, the team was able to probe which anti-cancer drugs were most effective at dealing damage to a simulated cancerous tumour.