Computational biology

Flagship leader

Modern biology and medicine are data driven. Fueled by the remarkable advances in experimental techniques, we are now able to identify the genetic blueprint of organisms in its entirety, elucidate the three dimensional structures of biomolecules at atomic resolution, interpret the emergence of complexity  through  the  study  of interactions of biological objects working in concert, and integrate microscopic and macroscopic information to explain the whole of the life processes in terms of a complex interplay of biological organisms. The potential significance and impact of analysing and interpreting biological  data  of  various  descriptions  has created an important cross-disciplinary area of modern science, Computational Biology, where computational and biological sciences converge.

This research flagship synergizes the Faculty's excellent expertise in Artificial Intelligence, Machine Learning, and Optimization, amongst others, to address major computational challenges that arise in biological data. The research in this flagship is organized into three complementary research themes,  catering  to  research  activities at the levels of:

  1. Macromolecules ( DNA , RNA and Proteins etc.)
  2. Systems (Regulatory, Metabolic etc.), and;
  3. Ecology (self-organization, simulation and modelling)

Flagship themes

Macromolecular informatics

Protein structural bioinformatics

Proteins are macromolecules of life responsible for almost all biological and cellular activities in living organism. Research in this area focuses on challenging problems at the forefront of computational research in structural biology, supporting the elucidation of the three-dimensional structures  of  proteins  and  the discovery of principles of protein architecture, function and evolution. This research area is directly aligned with Monash's key strength in protein structural biology and crystallography. Monash university provides excellent and extensive resources and infrastructure to support  a  number  of  research  activities in this area. These include: Clive and Vera Ramaciotti Centre for Structural Cryo-Electron Microscopy; Grollo Ruzzene Foundation Centre for Protein Structure, ARC Centre of Excellence in Structural and Functional Microbial Genomics, and the Australian Synchrotron.

Genomics of Disease

The cost-effective next generation nucleotide sequencing technologies provide researchers with genomic data with remarkable precision, at unprecedented rates. For the first time in human history, this has resulted in an ability to investigate at high resolution the molecular mechanisms driving diseases  such  as  cancer  among other heritable diseases. The focus of this area will be to identify the mutations and aberrations in the host genome that drive the disease manifestation and the underlying evolutionary mechanisms responsible for the progression, metastasis (in case of Cancer) and relapse during  the  life  span  of the disease.

Computational Systems Biology

Systems Biology Graphical Notation ( SBGN )

SBGN languages are intended to foster the efficient storage, exchange and reuse of information about signalling pathways, metabolic networks, and gene regulatory networks amongst communities of biochemists, biologists, and theoreticians. The simplicity of SBGN syntax and semantics makes SBGN maps suitable  for  use  at  the high school level. We are currently focusing on automatically producing SBGN maps for pathways in the KEGG repository. We plan to look at extending this work to other databases.

Computational Ecology

Computational Ecology employs computational and mathematical models to study how organisms interact with each other and with the inanimate environment, and how these interactions shape organisms, populations and the environment. Computational and mathematical models are an indispensable component of  ecological  studies,  as  ecological experiments are often impossible to conduct due to the size and complexity of the systems involved and due to the timescales being studied. Even where experiments are possible, they are usually difficult to conduct and hard to reproduce, so that virtual computational  experiments  provide  valuable  guidance for the design of real experiments. Some of the core techniques that we employ and develop further are: agent-based simulation, cellular automata, stochastic modelling, probabilistic graphical models, specifically Bayesian networks.

Our work currently focuses on:

  • Understanding social insects from the collective behaviour of their colonies to their evolutionary adaptation
  • Knowledge engineering for ecological risk assessment
  • Complexity in landscape ecology
  • Epidemiological simulation
  • Evolution theory

Main collaborations

Collaborating Organizations Members
BC Cancer Agency, Canada
  • Sam Aparicio
  • Sohrab Shah
Huck Institute for Genomics, Proteomics and Bioinformatics, Penn State, USA
  • Arthur Lesk
Indian Institute of Technology, Bio Systems Lab, Chemical Engineering Department
  • Pramod Wangikar
  • Sandeep Gaudana
Monash Faculty of Medicine, Nursing and Health Sciences
  • Ashley Buckle
  • Bosco Ho
  • Gail Risbridger
  • James Whisstock
  • Jiangning Song
  • Steve Bottomley
Systems Biology Institute, Japan
  • Hiraoki Kitano
  • Samik Ghosh
University of Melbourne, School of Population and Global Health
  • Emma McBryde
  • James McCaw
  • Jodie McVernon

Showcase projects

Inchman

Inchman is a system for spatially-resolved stochastic simulations on general-purpose graphics processing units ( GPGPU s) with widespread applications in computational biology and molecular chemistry.

On the mesoscopic level, Inchman implements the Gillespie method, which is widely used for spatially extended bio-chemical reactions.It provides one of the fastest implementations available, which has, for example, enabled us to, for the first time, build an ab-initio stochastic simulation of the Oregonator model for the Belousov-Zhabotinsky ( BZ ) reaction.

Inchman extends to the microscopic level with a very fast individual-based simulation of generalised reaction-advection-diffusion systems. This makes it suitable for modelling and simulation more complex collective systems, such as the organization of socially living organisms.

Self-organized Collective Decision Making in Dynamic Environments

Self–organisation is a fundamental mechanism used in nature to achieve flexible and adaptive behaviour in unpredictable environments.

A paradigmatic example of self-organised social groups is an ant colony, whose strikingly organised and seemingly purposeful behaviour at the group level is coordinated without any central "master plan" or leader.

Complex behaviour at the colony level emerges from simple interactions between myriads of individuals that only process local information. Similar forms of self-organised collective behaviour are found on all levels of size and complexity, from bacteria colonies via schools of fish, flocks of birds and herds of cattle to the social behaviour of human groups.

Our work is investigates specifically which organisational principles allow such behaviour to adapt efficiently to changing environments.

This work relies on stochastic modelling techniques and high performance simulation methods developed in a sister project.

Image © Alex Wild 2013

Pollination in a new climate

Agent-based simulations enable time- and cost-effective evaluations of complex interactions between different real-world entities. Recent investigations on pollinator-plant interactions show that the learnt flower preferences of important pollinators like bees is dependent upon both flower temperature, and regional ambient temperatures. This shows that local and global changes in climatic conditions may directly influence how certain plants are pollinated. This project is producing computer simulations to reveal how climate change may directly influence flower evolution in the future, and how the management of  environmentally and economically important plants can be modelled to inform reliable decision making about this important resource.

Modelling Infectious Diseases

Infectious diseases pose continuing and emerging challenges, with new strains of resistant diseases and novel varieties finding new means of transmission. Policy makers require models which can deliver improved predictions for assessing the value and consequences of public health interventions. This project develops mathematical models, agent-based simulations and Bayesian network models to advance our understanding of the epidemiology of infectious diseases.

Multiple Structural Alignment of Proteins

MUSTANG is a popular algorithm to simultaneously align multiple three-dimensional structures of protein. Among its many practical uses is the characterization of protein conformational change (e.g. evolutionarily conserved core residues) and production of multiple alternative search models for X-ray crystal structure determination from diffraction data using Molecular Replacement ( MR ).

DriverNet: uncovering the impact of somatic driver mutations on transcriptional networks in cancer

Simultaneous interrogation of tumor genomes and transcriptomes is underway in unprecedented global efforts. Yet, despite the essential need to separate driver mutations modulating gene expression networks from transcriptionally inert passenger mutations, robust computational methods to ascertain the impact of individual mutations on transcriptional networks are underdeveloped. We introduce a novel computational framework, DriverNet, to identify likely driver mutations by virtue of their effect on mRNA expression networks. Application to four cancer datasets reveals the prevalence of rare candidate driver mutations associated with disrupted transcriptional networks and a simultaneous modulation of oncogenic and metabolic networks, induced by copy number co-modification of adjacent oncogenic and metabolic drivers. DriverNet is available on Bioconductor.

PhD projects

Project title Supervisors
Simulating bee foraging: how behavioural diversity in bees interacts with environmental conditions.
  • Adrian Dyer ( RMIT )
  • Alan Dorin
Using co-evolution to predict protein-protein interactions
  • Geoff Webb
Orthogonal Layout of Biological Pathways in Garuda
  • Hiroaki Kitano (Systems Biology Institute, Tokyo)
  • Kim Marriott
  • Michael Wybrow
  • Samik Ghosh (Systems Biology Institute, Tokyo)
Computational and mathematical modelling of self organization in biological systems.
  • Bernd Meyer
  • Matthias Vigelius
Scientific visualization and analysis of simulated epidemiological data.
  • Alan Dorin
  • Kevin Korb
  • Nic Geard
Large scale modeling and inference of gene regulatory networks
  • Madhu Chetty
  • Rob Evans (University of Melbourne)
  • Terry Caelli (NICTA)
Computational Models for Identifying Driver Mutations in Cancer
  • Gail Risbridger
  • Reza Haffari
Next-generation protein structural comparison using information theory.