Project Description

A/Prof. James M. Hogan

Queensland University of Technology

How to make a really bad visualisation

Bio-data analysis and visualisation

Thursday 5 July 2018

James M. Hogan is an Associate Professor of Computer Science at the Queensland University of Technology, where he leads projects in bioinformatics visualisation and alignment-free sequence analysis. He received a BSc (Hons) from the University of Queensland in mathematics and a PhD in computer science from QUT, working on large scale randomly connected neural networks. His bioinformatics work has included machine learning methods for promoter and binding site detection, k-mer and Bloom Filter based approaches for sequence comparison, and visualisation of transcriptional regulatory networks and large scale sequence collections. He currently leads an ARC funded project in collaboration with CSIRO and Microsoft Research exploring the Visual Analytics of Next Generation Sequencing data, with a particular focus on extremes of scale and complexity.

The past decade has seen a dramatic rise in the scale of the data sets available in molecular biology and clinical practice. While the particular characteristics of these data inevitably change with the technology which generate them, as yet there is no sign at all that the revolutions in sequencing and related fields are coming to an end. It is now accepted wisdom that visualisation has a considerable role to play in the initial exploration and ongoing analysis of these data sets, and in the successful communication of the scientific findings which result. However, good visualisation is a great deal harder than it looks, and so we are not even going to try :)

In this talk, we will focus on a small number of example data sets—some from molecular biology, some from other disciplines—and teach you some of the tricks and oversights and misapprehensions which can make a visualisation truly terrible. Sometimes our mistakes can go virtually unnoticed, but at other times they may undermine the credibility of our work, masking findings that the world should really know about, or enhancing our perception of a significance that might not really be there.

As we learn to do things badly we might just realise how to do some of them well.

Not available.