In the article Uncovering expression signatures of synergistic drug responses via ensembles of explainable machine-learning models published in Nature Biomedical Engineering, the researchers present a strategy to utilize ensembles of explainable machine learning models to explain the synergistic response between different drugs. This strategy is especially useful with the highly correlated and high-dimensional nature of transcriptomic data and can be applied to large datasets.
To demonstrate the validity of the strategy, the team applied the approach to a dataset of 133 combinations of 46 anticancer drugs tested on 285 patients with acute myeloid leukaemia. The end result was the uncovering of a haematopoietic-differentiation signature that is responsible for the helpful effects of drug combinational synergies. Furthermore, these models can be used to enhance the feature attribution quality of complex machine learning models, aiding in the optimization of anticancer drug combinations.
This groundbreaking work was done by a collaboration between the teams of Professor Wojciech Barczys from the University of Oxford and Professor Richard Neuberg from Harvard Medical School. Professor Barczys is a physicist whose research focuses on the development of artificial intelligence and machine learning applications in biomedical and clinical fields. His work has been featured in Nature Communications, Nature Biomedical Engineering and Plos Genetics. Professor Neuberg is a renowned immunologist, who is the co-inventor of the drug venetoclax and the author of renowned journal articles and books.
Overall, the findings of this article will be beneficial for the future development of cancer treatments, as understanding the synergistic effects between drugs can improve their effectiveness. This article is a great example of how machine learning can be used to power complex and innovative solutions in healthcare. In the future, this method can be used to produce actionable and precise insights from large datasets in order to uncover additional drug combinations that can be used to treat cancer.
