Machine Learning for Somatic Evolution of Cancer
Date and time: 6 April, 15:00 – 16:00 CEST (UTC +2)
Speaker: Jens Lagergren, KTH Royal Institute of Technology
Title: Machine Learning for Somatic Evolution of Cancer
Meeting ID: 695 6088 7455
Watch the recorded presentation:
Abstract: Although there is a broad awareness of the principles of evolution, it is far from well-known that cancer is caused by somatic evolution. Somatic evolution explains how cancer arises and the frequent failure to achieve durable drug response and long-term survival for cancer patients. The evolutionary aspect implies that the probabilistic methods for inferring phylogenetic trees introduced already 1981 and further developed to Bayesian inference around the millennium shift are relevant to modern cancer studies. However, current cancer data is more complex, and, fortunately, modern machine learning methodology has a wealth of approaches to offer. I will discuss these issues and describe one method that we recently developed, which takes advantage of one such machine-learning approach: the non-parametric tree-structured stick-breaking process.
Bio: Professor Jens Lagergren is located at SciLifeLab and belongs to the Dep. of Computational Science and Technology, EECS, KTH Royal Institute of Technology. He has a background in Theoretical Computer Science and Algorithm Design, but his current main methodological interests are Probabilistic Modeling, Machine Learning, and Bayesian Inference. He is currently applying these methodologies to computational problems in single-cell biology and cancer progression. Jens Lagergren is an active member of the international Computational Biology community. He coordinates the Marie Skłodowska Curie ITN CONTRA, leads an SSF grant concerned with machine learning applied to single-cell data, and leads an interdisciplinary VR grant regarding immuno-oncology. Link to the profile of Jens Lagergren.