Bioinformatics Seminars

Bioinformatics Seminar

Time: 10:45am Tuesdays.
Level 7 Seminar Room 2, WEHI1

5 September 2017

Exploring Tumor Evolution

Lachlan McIntosh
WEHI Bioinformatics

We would like to deconvolve the average signal (copy numbers/structural variations/snvs) of multiple heterogeneous samples of populations of tumor cells to their underlying genomes and corresponding cellular fractions. Then we would like to take these discrete snapshots and build a picture of the evolution of cancer. In particular, we would like to build mathematical models that match our understanding of biological mechanisms so that we can systematically study these mechanisms. This task is made difficult by the unique and varying properties of data types collected from various experiments and so we must take advantage of natural constraints learnt by developing familiarity with the data.

In this talk I will talk about deconvolution, working with noisy data (GC correction for highly aneuploid data, dye bias in snp arrays, and more), and then focus on a particular aspect of cancer evolution that is surprisingly not well studied in the literature; the mechanism of "genome doubling" (where all the dna within the cell doubles at the same point in time).

Search past seminars