An interdisciplinary research team of biomedical and computer scientists from the Swiss Federal Institute of Technology in Zurich (ETH), the University of Zurich (UZH) and the University Hospital Zurich (USZ) has developed a unique approach that may form the basis for more targeted and personalized treatments for diseases such as cancer. To this end, they have combined machine learning with the mathematical theory of optimal transport (OT).
“The diversity within a group of cells greatly influences their sensitivity or resistance to changes,” explains Gunnar Rätsch, Professor of Biomedical Informatics at the ETH and USZ, in an ETH article. “Instead of basing our understanding on the average response of a cell group, our method can precisely describe – and even predict – how each cell reacts to disturbances like those from a drug.” In their recently published study, the researchers show that their method works not just for cancer cells but also for other pathogenic cells – for example, in the case of autoimmune disease lupus erythematosus.
The new prediction method is called CellOT and, according to the article, is the first approach to use optimal transport and machine learning to predict the perturbation responses of cells from new samples – including cells whose reactions have not yet been measured in the laboratory. “Established OT methods do not allow for out-of-sample or out-of-measurement predictions,” says Charlotte Bunne, one of the study’s three lead authors. “But that’s exactly what CellOT can do.” ce/mm
Related news
Meet with an expansion expert
Our services are free of charge and include:
- Introduction to key contacts in industry, academia, and government
- Advice on regulatory framework, taxes, labor, market, and setting up a company
- Custom-made fact-finding visits, including office and co-working space