Dr Luke Mathieson
School of Elect Engineering and Computer Science
- Phone:(02) 40420832
It all adds up
Pairing powerful computer technology with applied mathematics, Dr Luke Mathieson's data analyses are a welcome, revolutionary step toward personalised patient care.
Crunching numbers isn't for everyone. For Dr Luke Mathieson though, it's a gratifying challenge. The enthusiastic computer scientist is both a calculator and interpreter at the University of Newcastle's Centre for Information-Based Medicine (CIBM), approaching complex questions from novel algorithmic perspectives.
"I extract meaning from really big data sets," he explains.
"These sets are basically a bunch of figures in a file that tell you how molecular components interact."
"My job is to build networks to examine the links between them, with the eventual aim of pinpointing disease-related subtypes."
Consequently aiding the identification of key biomarkers capable of tracking – and perhaps blocking – the progression of cancer and neurodegeneration, Luke's studies are rapidly transforming multiple medical fields. Drawing together a number of clinical and academic disciplines, they're also an exciting new frontier where 'bench to bedside' research is concerned.
"The work I do is largely theoretical, but it has some very practical, 'real world' applications too," he affirms.
"We're looking to develop patient-tailored treatments for a host of conditions that are influenced by genetic aspects, for example, such as melanoma and schizophrenia."
Luke began his research career with an Honours Degree at the University of Newcastle in 2004. Focusing on a "very small slice" of statistical database security, the 12-month investigation looked to balance a universal trade-off between access and privacy.
"My project was essentially about merging or masking sensitive information in such a way that ensures the databases you get at the end are still useable."
"Unfortunately though, I found that even if you take a very simple theoretical model, you can't prevent attacks without rendering them completely useless."
"Protecting files is therefore an almost impossible aim – if they're open to legitimate use, they're similarly open to illegitimate misuse."
Out of the 'too hard' basket
Decisively switching areas after completing his undergraduate program at the end of 2004, Luke opted to undertake several assignments on complexity theory with Professor Pablo Moscato. Equally impressed by the University's invention of an offshoot, known as parameritised complexity, Luke found a home for his research ambitions at the CIBM for the next two years.
"Our main goal was to find structure in data," he shares.
"The patterns we discovered were later used to produce sensible calculations."
Duly inspiring a global reexamination of problems once thought to be unsolvable, these studies led into Luke's PhD work at England's Durham University in 2007.
"During this time I developed a series of algorithms for some related complications in graph-editing," he reveals.
"Graphs are the mathematical term for networks."
"If you draw them on paper, which we often do, they're a bunch of circles and lines."
"The circles model things, like proteins or genes, and the lines represent their interactions."
Though comically admitting his candidature "probably sounds incredibly boring" to most people, Luke concedes he unearthed a whole set of interesting problems that are "really hard but not impossible" to crack during the three-year probe.
"With each graph, we have a small set of properties that we're wanting to satisfy," Luke describes.
"So vertices and edges have to be added and/or removed to make them fit."
"These graphs are then given to computers as inputs, telling them to do certain steps and generate a 'hopefully' predetermined product or result."
"Some are easy to produce algorithms for, meaning we can get computers to solve the problems correctly and quickly without using many resources, and others are not so easy."
Old and new
Luke returned to Australia after receiving his award, collaborating with Pablo for another year before moving to Macquarie University in 2011. The esteemed mathematician continued to worked on parameritised complexity throughout the three-year posting, combining it with dynamic graph theory to model a handful of mobile ad hoc networks.
"These are the sort you get from our phones and cars and computers talking to each other," he enlightens.
"Each has a circle, but the lines and edges and vertices that connect them are constantly moving."
"This causes quite dramatic changes in the practice of our graphs."
Seeking to use maths in fresh and fascinating ways, Luke also undertook some "interesting guess work" to characterise pairwise relations between the devices.
"It looks very ugly to the public," he admits.
"All of a sudden some of the old algorithms that you used to use don't work anymore, so you're forced to come up with new ones."
Twice the trouble, twice the fun
Luke expanded this niche after returning to Newcastle in 2015, developing a practice called 'network alignment.' This time working with two or more networks, he is looking to detect a series of intricate sub-networks common in multiple species.
"What we want to do is line them up so that all the circles, edges and vertices map to each other," he states.
"If they line up well, this tells us that they share a very similar structure."
"For example, if you take the protein-protein interactions of a well-studied organism, like a nematode worm, and line them up with the human protein-protein interactions, we can figure out the basic biological mechanisms each is controlling."
Disclosing his long-term goal of automating this process "as much as possible" for personalised drug therapy, Luke advises that the same can be said for healthy and unhealthy human networks.
"If we can identify the proteins that are doing funny things, we can start to create a targeted approach," he suggests.
"This would be much more effective than chemotherapy, for instance, which kills the patient almost as quickly as it kills the cancer."
A master at multitasking, Luke is also currently exploring 'super networks.' Employed to do so under an Australian Research Council Discovery Project Grant, he is applying innovative mathematical techniques to build alignments that satisfy a number of different criteria.
"These networks will tell us how protein and gene data interact, or how a disease progresses," he insists.
"But it's not a simple task – we're hoping to make them information-rich and this requires more than just nodes and lines."
"Super networks are really a whole bunch of individual networks put together."
"Each is about 200GB, which is seriously pushing the limits of what you can do with a computer."
With parallel computation, a form of computation in which many calculations are carried out simultaneously, making it easier and quicker than ever before to "do something sensible" with such datasets, Luke and his colleagues are well on their way to achieving what was once deemed impossible.
"Large problems can often be divided into smaller ones, which are then solved at the same time," he elucidates.
"We're finally getting to the point where we can disentangle them and extract meaningful biological and medical information."
- Doctor of Philosophy, University of Durham
- Bachelor of Science, University of Newcastle
- Bachelor of Computer Science, University of Newcastle
- Bachelor of Computer Science (Honours), University of Newcastle
- Computational Complexity
- Data Structures
- Discrete Mathematics
- Graph Theory
- Object Oriented Programming
- Parameterized Complexity
- Programming Language Theory
- Software Engineering
- Theory of Computation
- English (Mother)
Fields of Research
|080201||Analysis of Algorithms and Complexity||50|
|080202||Applied Discrete Mathematics||30|
|Title||Organisation / Department|
|Research Associate||University of Newcastle
School of Elect Engineering and Computer Science
|Dates||Title||Organisation / Department|
|1/01/2014 - 1/09/2014||Adjunct Lecturer||Macquarie University
Department of Computing
|1/05/2011 - 1/12/2013||Postdoctoral Research Fellow||Macquarie University
Department of Computing
|1/01/2010 - 1/05/2011||Postdoctoral Research Associate||University of Newcastle
Centre for Information-Based Medicine
For publications that are currently unpublished or in-press, details are shown in italics.
Journal article (9 outputs)
Mathieson L, 'Synergies in critical reflective practice and science: Science as reflection and reflection as science', Journal of University Teaching and Learning Practice, 13 (2016)
Â© 2016, University of Wollongong. All rights reserved.The conceptions of reflective practice in education have their roots at least partly in the work of Dewey, who describes ref... [more]
Â© 2016, University of Wollongong. All rights reserved.The conceptions of reflective practice in education have their roots at least partly in the work of Dewey, who describes reflection as Â¿the active, persistent, and careful consideration of any belief or supposed form of knowledge in the light of the grounds that support it and the further conclusions to which it tendsÂ¿ (Dewey 1933, p.9). This conception of reflection has carried on into more-focused efforts to describe critical reflection as a tool for improving professional practice (where academic and educational practice is the particular interest of this study); Â¿Â¿ some puzzling or troubling or interesting phenomenonÂ¿ allows the practitioner to access Â¿the understandings which have been implicit in his action, understandings which he surfaces, criticizes, restructures, and embodies in further actionÂ¿ (SchÃ¶n 1983, p. 50). Both of these descriptions embody a central idea of critical reflective practice: that the examination of practice involves the divination (in a rational, critical sense) of order and perhaps meaning from the facts at hand (which, in turn, are brought to light by the events that occur as the results of implementation of theory). As part of a lecture series, Gottlieb defined science as Â¿an intellectual activity carried out by humans to understand the structure and functions of the world in which they liveÂ¿ (Gottlieb 1997). While science and critical reflective practice attempt to build models about different parts of our world Â¿ the natural world and the world of professional (educational) practice respectively Â¿ both embody certain underlying aims and methodologies. Indeed, it is striking that in these definitions the simple replacement of the terminology of reflective practice with the terminology of science (or vice versa) leads to a perfectly comprehensible definition of either.It is this confluence that this paper studies, building from two separate foundations, critical reflective practice and science. Via their models and exemplars of their Â¿models-in-practiceÂ¿ Â¿ action research and the scientific method Â¿ the paper forms a bridge between two empirical practices. We contend that the ability to do this is no accident, but stems from a deeper substrate that they have in common: empirical epistemology, as expressed in post-enlightenment models of the development of reliable knowledge.
Konyagin SV, Luca F, Mans B, Mathieson L, Sha M, Shparlinski IE, 'Functional graphs of polynomials over finite fields', Journal of Combinatorial Theory, Series B, 116 87-122 (2016)
Frati F, Gaspers S, Gudmundsson J, Mathieson L, 'Augmenting Graphs to Minimize the Diameter', Algorithmica, 72 995-1010 (2015)
Mathieson L, Szeider S, 'Editing graphs to satisfy degree constraints: A parameterized approach', Journal of Computer and System Sciences, 78 179-191 (2012) [C1]
Arefin AS, Mathieson L, Johnstone DM, Berretta RE, Moscato PA, 'Unveiling clusters of RNA transcript pairs associated with markers of Alzheimer's disease progression', PLOS One, 7 1-25 (2012) [C1]
Mellor D, Prieto-Rodriguez E, Mathieson L, Moscato PA, 'A kernelisation approach for multiple d-hitting set and its application in optimal multi-drug therapeutic combinations', Plos One, 5 1-13 (2010) [C1]
Rizzi R, Mahata P, Mathieson L, Moscato PA, 'Hierarchical clustering using the arithmetic-harmonic cut: Complexity and experiments', Plos One, 5 1-8 (2010) [C1]
|Show 6 more journal articles|
Conference (8 outputs)
Mathieson L, Mans B, 'On the Treewidth of Dynamic Graphs', Computing and Combinatorics (2013)
Mathieson L, Frati F, Gaspers S, Gudmundsson J, 'Augmenting Graphs to Minimize the Diameter', Algorithms and Computation (2013)
Arefin AS, Inostroza-Ponta M, Mathieson L, Berretta RE, Moscato PA, 'Clustering nodes in large-scale biological networks using external memory algorithms', Algorithms and Architectures for Parallel Processing (2011) [E1]
Mellor D, Prieto-Rodriguez E, Mathieson L, Moscato PA, 'Uncovering combinations: Using graph theory to find multi-drug therapies', Biomarker Discovery Conference. Poster Program (2010) [E3]
Moscato PA, Mathieson L, Mendes ADS, Berretta RE, 'The Electronic Primaries: Predicting the U.S. Presidency Using Feature Selection with Safe Data Reduction', Proceedings of the twenty eighth Australasian Computer Science Conference (ACSC 2005) Newcastle, Australia, January, 2005 (2005) [E1]
Mathieson L, King T, Brankovic L, '2-Compromise: usability in 1-dimensional statistical database', Proceedings from Fifteenth Australasian Workshop on Combinatorial Algorithms (2004) [E1]
Mathieson L, Prieto-Rodriguez E, Shaw PE, 'Packing edge disjoint triangles: a parameterized view', Parameterized and Exact Computation (2004) [E1]
|Show 5 more conferences|
Thesis / Dissertation (1 outputs)
|2010||Mathieson L, The Parameterized Complexity of Degree Constrained Editing Problems, University of Durham (2010)|
Grants and Funding
|Number of grants||1|
Click on a grant title below to expand the full details for that specific grant.
20111 grants / $50,000
Computational prediction and functional clarification of novel drug combination strategies for the treatment of brain tumours$50,000
Funding body: Maitland Cancer Appeal Committee
|Funding body||Maitland Cancer Appeal Committee|
|Project Team||Professor Pablo Moscato, Dr DAN Johnstone, Associate Professor Regina Berretta, Doctor Luke Mathieson, Professor Manuel Graeber, Doctor Jennette Sakoff|
|Type Of Funding||Donation - Aust Non Government|
Dr Luke Mathieson
School of Elect Engineering and Computer Science
Faculty of Engineering and Built Environment
|Room||Level 3 Pod|
|Location||1 Kookaburra Circuit, John Hunter Hospital