top of page
THE PROJECT
Mutations in a cell’s genome can cause it to divide uncontrollably and result in cancer. All cells in a tumour are descendants of that original ancestor cell and they can acquire new mutations along the way. Single cell genome sequencing and computational inference reconstruct the evolutionary relationships between cells to better understand how tumours develop and how to treat them.
THE DESIGN
variation
The first key challenge was to visualize genome alterations of individual cells together with the evolutionary tree structure inferred from this information. CellScape offers a coordinated tree and heatmap visualization to help scientists explore the specific mutations that are prevalent across all cells or those that have emerged in only a subpopulation of cells.
variation over time
It's also informative to visualize how cancer genomes evolve over time. Inspired by hand-drawn Muller diagrams, TimeScape reveals temporal changes in the prevalence of genetically distinct cells observed in single cell sequencing from tumour samples over a patient's course of treatment. Such a view helps scientists understand how tumours evolve in particular after therapeutic interventions like chemotherapy.
variation across anatomical sites
In some cases, cancers can spread to different parts of the body and it's important to understand the genomic changes that drive that spread. Comparing the genomic landscape of tumour cells in different locations leads to insight into the mechanisms of metastasis as well as potential targets for treatment.
MY CONTRIBUTIONS
DEFINE
the problem
This project is the thesis work of Maia Smith (UBC Bioinformatics Graduate Program) whom I co-supervised together with Sohrab Shah. I helped to define the initial project goals and scope.
DESIGN
the prototypes
Throughout the design process, I enjoyed collaborating closely with Maia and offered advice on both the technical detail and high-level approach.
WANT TO LEARN MORE?
E-scape was published as a part of Smith et al. 2017 in Nature Methods and the software packages are available open-source from Bioconductor: CellScape, TimeScape, and MapScape.
also see
bottom of page