Research Spotlight (PFNS/UWE/UOB): Harnessing adaptive novelty for automated generation of cancer treatments

https://www.sciencedirect.com/science/article/pii/S0303264720301684

Our team members from the University of Novi Sad, University of the West of England, and University of Bristol have just released a paper on EVO NANO’s simulation capabilities.

For the modelling, they use a Python-based platform called Mesa – which allows cancer cells, cancer stem cells, and nano-agents to be modelled in 2D. The nano-agents are varied and given values based on how well they kill or inhibit the different cancer cell types. This value then allows the model to select the top 5% and remove the bottom 5% performing nano-agents, and mutates the middle 5%. In this way, the nano-agents evolve and improve in fitness far more rapidly than a model that has no evolution.

From the author, Igor Balaz:
We showed that this strategy is especially efficient when dealing with the emergence of tumour resistance. In such scenarios, nano-agents are able to continuously track down newly emerging tumour mutations and keeps them under control.