CCBIO Special Seminar – Iain G. Johnston
Join us for an interesting seminar by Associate Professor Iain Johnston, titled "HyperTraPS: Learning pathways of disease (and cancer) progression from data."
Algorithm to predict disease course
Mathematician Iain Johnston has recently moved to Bergen to work at the Department of Mathematics, University of Bergen. Together with researchers from the Imperial College London, he has developed a new tool, "HyperTraPS" (hypercubic transition path sampling), with an algorithm that can uncover not only overall structures in how diseases develop, but also, based on a given set of symptoms, with a high probability predict the next stage of the disease course for each patient.
HyperTraPS can efficiently learn progression pathways from cross-sectional, longitudinal, or phylogenetically linked data, readily distinguishing multiple competing pathways, and identify the most parsimonious mechanisms underlying given observations.
Applied on ovarian cancer data
The tool has among other been applied to data from thousands of ovarian cancer patients. The analyzes show different disease pathways that largely depend on which mutation came first. The study was recently published in Cell Systems. You can find a summary of the article "HyperTraPS: Inferring Probabilistic Patterns of Trait Acquisition in Evolutionary and Disease Progression Pathways" here.
Also see interview with Iain Johnston in Dagens Medisin (in Norwegian).
When: January 31st 2020 at 15.15-16.00
Where: Auditorium H113, Haukeland University Hospital, main building (Sentralblokka), 1st floor.
Registration: not necessary, open seminar.
Who: Open to all who are interested.