MMSN -- About Us

The primary goal of the Network is the building of a new and powerful e-Science tool that will ensure that Australian scientists are exceptionally well equipped to push the global leading edge of any research that depends on a knowledge of structure at a molecular level. The MMSN will collaboratively develop two closely related internet network services to foster and advance molecular and materials structure e-Science and its diverse application and utilisation in the broader scientific community.

A remote instrument access network will make available from a user's desktop sophisticated instrumentation at various laboratories around the country, including instruments at the Replacement Research Reactor and the Australian Synchrotron. The desktop instrumentation access will be partnered with a national database and visualisation service supporting collaborative interactions.

The establishment of these network services and the 21st century collaboratory (Wulf, Science, 261, 854 (1993)) they will constitute will significantly enhance research endeavours in chemistry, materials science, biology and computer science, and will catalyse the formation of new linkages between these sciences. Input from the user community represented by the diverse membership of the MMSN will ensure the collaboratory has real word functionality and is user friendly.

The MMSN participants include internationally distinguished Australian scientists, leading members of the Australian Nuclear Science and Technology Organisation (ANSTO; responsible for the Replacement Research Reactor project), the Australian Synchrotron Research Program (ASRP), members of the National Scientific Advisory Committee for the Australian Sunchrotron, and has the support of the Australian Synchrotron Project: "The Australian Synchrotron Project recognises the potential of automation and remote access programs in providing the most efficient and effective use of beamlines by Australia's distributed research community, and is interested in the development of such expertise." The proposed network also has the support of The Australian Institute of Nuclear Science and Engineering (AINSE).

Context

Knowledge of the three dimensional arrangement of atoms, and the chemical identity of those atoms, in molecules and solid state materials is crucial to understanding their behaviour and harnessing their potential in real-world applications. Such knowledge can deliver profound scientific, technological, social and economic rewards.

Understanding the biological process that support and shape life and the disease and degenerative process that may threaten life, depends increasingly on a detailed knowledge of bio-molecular structure; protein structure, virus structure, DNA complex structure, and enzyme structures such as that of the SARS protease. Rational medicinal drug design, which for example produced the first anti-flu drug Relenza (developed by CSIRO) and the HIV protease inhibitors, would be impossible without a knowledge of molecular structure.

The determination and rationalization of the relatively small atomic structures comprising micro-magnets, microporous and mesoporous materials, hydrogen storage materials, novel metal oxides, ceramics, superconductors, minerals, 'smart' materials, piezoelectric materials, magneto- and electro-rheostatic materials, photonic devices, information storage devices, molecular switches and sensors, biomimetic materials, and pharmaceutical materials is crucial to their development and utilisation.

The past 100 years or so has seen the evolution of a very powerful set of analytical tools for the determination of molecular and solid state structure. Some of these tools can be accommodated in conventional laboratories (so called 'home labs'), whereas increasingly more powerful probes require major infrastructure facilities such as the Australian Synchrotron being built in Victoria and the Replacement Research Reactor being built in New South Wales. Techniques used both in conventional labs and at major facilities include:

  • X-ray diffraction from single crystals; this is the most widely used technique for accurately and precisely determining structure at atomic resolution. Sophisticated procedures make the growth of crystals of materials of interest increasingly routine, in fact robots are now being used. Crystals are comprised of regular arrangements of atoms (often in discrete molecules) and illuminating a crystal with X-ray light results in the light being scattered in all directions, from the electrons surrounding the atoms, forming a characteristic diffraction pattern. The diffraction process is determined by the nature of the atoms (ie the number of electrons around each atom) and their three dimensional arrangement. Measuring the diffraction pattern can then provide the atomic structure.
  • X-ray diffraction from powders; when single crystals cannot be grown structural information may nonetheless be gleaned from grinding the crystalline material of interest into a micro-crystalline powder. X-ray diffraction from the powder can be sufficiently characteristic to enable a complete structure determination. The method is currently less effective than using single crystal diffraction and the development of structure determination methods using powder diffraction is an area of very active research.
  • Small Angle and Wide Angle X-ray Scattering (SAXS/WAXS); provides a structural information from materials that do not have the highly ordered arrangements of component atoms found in crystals. The SAX technique gives size and shape structural information at scales of 500 to 0.1 nanometers (0.000000001 of a meter), whereas WAX provides information to atomic resolution.
  • X-ray absorption spectroscopy (XAS); measuring the absorption of X-rays (rather than scattering) as the light wavelength (energy) is varied reveals characteristic absorption maxima called absorption edges that occur when the light has sufficient energy to move an electron in the material to a higher 'energy level'. There is characteristic fine structure near the absorption edge as the light energy increases as the level of absorption declines. Extended X-ray Absorption Fine Structure (EXAFS) provides structural information such as atom to atom distances and the number of atoms 'bonding' to any metal atoms that may be present. X-ray Absorption Near Edge Structure (XANES) yields information about the geometrical arrangement of atoms and the oxidation state of atoms. X-ray absorption spectroscopy does not require crystalline material and liquid solutions are often studied. Biological cells can be studied directly, as can soil samples for environmental analysis. At longer wavelengths X-ray absorption spectroscopy can be used to study surfaces and thin films.
  • X-ray Photoelectron Spectroscopy (XPS); is an analytical technique that depends upon the measurement of the energies of photoelectrons that are ejected from atoms when they are irradiated by low energy X-ray light. Used to study surfaces, XPS has a number of attributes including a high (and variable) range of sensitivities to structures on the outermost surface, an ability to identify such structures chemically, and a reasonable capacity for elemental quantification and structure thickness.
  • X-ray microtomography is an imaging technique that uses computed tomography to construct high resolution 2 and 3 dimensional images of small objects. The high contrast and microtomography capabilities proposed for the Australian Synchrotron will be exploited to great effect in the areas of materials science, non-destructive testing and mineralogy.

Synchrotron sources provide light many orders of magnitude brighter than can be obtained in a conventional laboratory, and the wavelength of this light can be tuned to enable experiments and measurements otherwise impossible.

Neutron diffraction and scattering provides structural information that complementing that gained from the scattering and absorption of light. Neutrons interact with the nucleus of an atom, rather than the electrons surrounding the nucleus, and these interactions can locate atoms positions more accurately than can X-ray interactions with electrons, which may be distributed between atoms.

The powerful instrumentation used in structurally characterising molecules and solid state materials is inherently expensive to purchase or construct. This is obviously true for instruments at major facilities, and it is also true of uniquely capable instruments housed in conventional laboratories. Such equipment is inevitably too costly to replicate and maintain in multiple locations. Practitioners currently travel in teams to use instrumentation not locally available; this is both costly and time inefficient. Fortunately recent technological developments make automation and the provision of Internet access to remote instrumentation a realisable possibility.

The Molecular and Materials Structure Network seeks to exploit the emergence of e-Science and Grid computing, in building a collaborative network environment providing internet access to remote instrumentation. The same environment will host national database resources for the molecular and materials structure sciences, and provide powerful resources to interactively display, manipulate, analyse and discuss atomic structures (biological or chemical; discrete, polymeric, 'infinite lattice' or otherwise) across multiple monitors anywhere in the country.

The Grid is a concept that has evolved from cluster computing over the past decade, becoming particularly prominent as a potentially global revolution in computing following a publication in Nature by Ian Foster in December 2000. Essentially the Grid promises to seamlessly distribute and manage computations and data over participating computers that have unused computational or storage capacity. The underlying operating system or type or computer will be irrelevant, and the capacity of the Grid increases with the number of linked computer resources. The power, capacity and cost effectiveness offered by Grid computation and storage is being pursued particularly actively by the international particle physics community. The US is well advanced in developing its Science Grid (http://www.sciencegrid.org/), which is to link instrument facilities, data repositories and collaboratories, as is the UK and Europe.

There is increasing interest and investment in using Grid infrastructure for e-Science; science carried out through distributed collaborations hosted and enabled by the Internet. In the US e-Science is emerging with the development of the 'Collaboratory'. A Collaboratory is a 'meta-laboratory' that spans multiple geographical areas with collaborators interacting via electronic means. Collaboratories are intended to enable interactions between scientists in a given research area, promote cross-disciplinary collaborations, accelerate the development and dissemination of basic knowledge, and minimize the time-lag between discovery and application.

The UK and Europe are investing heavily in e-Science programs; the UK for instance has established a National e-Science Centre. Like Australia, the UK is currently building a new synchrotron, and the use of e-Science has been included in the planning for the facility. In the UK e-Science is seen as vital to the successful exploitation of its next generation of powerful scientific facilities, and to the UK's effective use of major facilities elsewhere

The molecular and materials structural sciences inherently provide an ideal partner for computer science, in exploring and establishing the benefits of e-Science and the Grid (see Buyya et al 2003 for a drug discovery example). For instance, each year chemical crystallography produces large quantities of at least three dimensional data in relatively small 'bundles' that represent molecular structures, and this data is particularly well suited to dissemination, visualisation, analysis and discussion over a an internet network. The same is true of biological structures, though the structure model is much larger and the computational and network demands much greater. Reflecting the rising importance of Grid computing for the structural sciences, the American Crystallography Association is focusing on Grid computing in its Advances in Computing Environments for Crystallography session. Similarly, the American Chemical Society has programmed a session entitled Research Collaboratories, Virtual laboratories and Grid Computing for its 2004 meeting.

International Progress

The US and Europe are investing heavily in the Grid and e-Science. The UK is embedding Grid based e-Science in the planning for its new synchrotron. The e-HTPX program aims to utilise the Grid in developing an e-Science resource for high throughput protein crystallography. Iniial implementation will be at the Daresbury synchrotron and at beamline 14 of the European Synchrotron Research Facility, with a view to being transferable to protein beamlines at Diamond. The intention is to implement Grid-based portals for protein crystallography, enabling remote access to all facilities over the internet. The program will extend and develop structure determination software to take advantage of low-cost, highly parallel Grid computing facilities so that feedback can be provided on the success, or otherwise, of phasing on the same timescale as data collection. Also being developed is a Grid-based application allowing the user to manage flow of data from the initial stages of target selection to the automated deposition of the final refined model in the public databases.

The e-HTPX project forms part of the European Union Bio-XHIT program which has recently been funded at 10.5 million Euros for 4 years. All stages of protein production, protein crystal growth, diffraction data, the determined structure and all meta-data will be recorded and stored in a data-Grid. Data collection and structure determination will be accelerated with robotics and secure remote access, and Grid computing will be used to automate and accelerate the structure determination.

A somewhat similar approach is being developed at the UK's National Crystallography Service for 'small molecule' structure determinations, which is based at the University of Southampton. Robotics are used to automate the data collection process, with a remote user being able to have input into the crystal selection process and data collection strategy. Shortly before a scheduled sample is about to be mounted the remote user is emailed notification, and the user can then start the sample screening and data collection process via Web browser access to the systems experiment tracking and control system interface (Laboratory Information Management System). The web interface delivers JPEG images from the intrumentation for the remote users evaluation. Software is being developed to automatically use the collected data to determine and refine the molecular structure, and then deposit the raw data and structure data in a publicly accessible data repository (part of the UK data Grid infrastructure).

In contrast to the multi-national collaborative approach to automation and remote access being taken in Europe, several essentially independent initiatives are approachig maturity in the US. The Stanford Synchrotron Radiation Laboratory developed a pioneering robotic system that has proved relatively inexpensive and robust (REF), and is developing a collaboratory remote access system. A similar program is underway at the National Synchrotorn Light Source at the Brookhaven National Laboratory, and likewise a remote access program is under consideration at the Advanced Light Source in Berkley. The South East Area Collaborative Access Team at the Advanced Photon Source of the Argonne National Laboratory is essentianlly mandated t provide remote access, and is planning to use Access Grid technology to deliver remote access.

Expected to be completed in 2006, the Spallation Neutron Source (SNS) is an accelerator-based neutron source being built in Oak Ridge, Tennessee, by the US Department of Energy. At a total cost of US$1.4 billion the SNS will provide the most intense pulsed neutron beams in the world for scientific research and industrial development. The facilities anticipated use of Grid techniques for data processing and storage is expected to significantly influence Grid development in the US, and hence the world. Remote access is being planned for this major facility, though remote user input will be limited and feedback to the user will likely use mininal graphics.

Australian Expertise

The electron microscopy community has pioneered research into the provision of remote access to structure determination instrumentation both nationally and internationally (https://telescience.ucsd.edu/). In Australia the Nanostructural Analysis Network Organisation (NANO) has expertise in the provision of remote instrument access. Instrumentation at the University of Queensland NANO node is already accessible through the Access Grid, and remote access is being developed at the University of Sydney node. The Director of the Nanostructural Analysis Network Organisation (NANO) is an MMSN participant, and NANO expertise in the provision of remote instrument access will be utilised in developing the MMSN network; the relationship will inevitably be synergistic.

Australia is fortunate in having four leading Grid Research groups; that of Prof. David Abramson at Monash University, Prof. Albert Zomaya, who holds the Cisco Internetworking Chair at Sydney University, Prof. Bernard Pailthorpe at Queensland University, and Dr Raj Buyya at Melbourne University. These researchers are internationally prominent, making substantial development contributions and leading major Grid computing conferences. Prof. Albert Zomaya is a distinguished expert in computer networks and parallel computing, and heads the Advanced Networking Research Group at Sydney University. Prof. David Abramson led the development of Nimrod-G which has become a standard Grid e-Science application for distributed parametric studies. Prof. Abramson and Dr Buyya subsequently applied Nimrod-G computing to search and analyse large molecular database content for drug discovery studies The recent inception of the GriddLeS project is of particular relevance to the MMSN in promising to simplify the utilisation of legacy software in a Grid environment. Dr Buyya is leading the Gridbus Project developing a grid economy-driven Data Grid and virtual organisation creating technologies for e-Science and e-Business applications. Prof. Bernard Pailthorpe is the Director of the Advanced Computational Modelling Centre at the University of Queensland, with expertise in modelling, visualisation and Grid computing. All four groups are participants in the Molecular and Materials Structure Network.

Australia has a well established pedigree in the molecular and materials structure sciences. The Braggs pioneered the use of X-ray diffraction for molecular structure determination, utilised today in successful rational drug design programs such as the development of the first anti-viral drug, Relenza, by Prof Peter Colman and Prof Jose Varghese. Prof David Winkler at the CSIRO Molecular Science Division has developed a powerful structure-activity analysis package called MOLSAR, for drug discovery research. Associate Professor Jenny Martin leads a protein structure group at Queensland University using rational drug design methods in the search for better pharmaceutical agents. Prof Michael Parker is one of the world leaders in the study of glutathione transferases, and membrane-associating proteins.

Prof Mark Spackman, ARC Professorial Fellow at the University of New England, is internationally a leading computational molecular structure researcher. In addition to seminal experimental and theoretical studies in the determination and analysis of electron distributions in molecules, central to their chemical and biological properties, his group has developed a powerful graphical representation of molecular interactions in the solid state (Hirshfeld Surface).

Prof Sydney Hall from the University of Western Australia is likewise a very eminent computational and theoretical structure researcher. Prof Hall developed a particularly powerful and flexible package, XTAL, for structure analysis using X-ray diffraction data. Prof. Hall, was also responsible for the design and development of the Self-defining Text Archive and Retrieval (STAR) file used widely in the structural sciences for representing, exchanging and verifying scientific data. Its discipline specific application in crystallography as the Crystallographic Information File (CIF) has become ubiquitous as the standard structural representation, data exchange and validation format. The CIF is used for the CSD, PDB and ICSD databases, and is used by the IUCr for structure validation. A CIF is now required by Journals for peer review of small molecule structures submitted for publication.

Prof Hyde, Dr Tomaso Aste and Dr Arthur Sakellariou from the Australian National University, have diverse interests in theoretical and experimental structural science, including the use of tomography for materials structure analysis.

Outline of Research Program

Remote Instrument Access
The MMSN will draw together national and international expertise in remote instrument access and automation in developing efficient use of Australia's molecular and materials structure instrumentation. The MMSN participants include representatives from major international facilities in the process of developing remote access programs. In Australia, the Nanostructural Analysis Network Organisation (NANO) has expertise in the provision of remote instrument access and is represented in the MMSN.

Developing a harmonised approach to providing remote access to various instruments encourages user-friendliness and re-usability. The same fundamental approach will be taken to providing remote access to both conventional laboratory and major facility instruments. The development program will not be constrained by the inflexibility of a 'one size fits all' approach, but will be adaptable to the characteristics and idiosyncracies of the individual instruments. Examples of 'conventional' instruments include the high performance chemical crystallography instrument to be installed at Sydney University in May 2004, the high brightness protein crystallography system at the University of Queensland and unique surface science instrumentation at the Universities of Queensland and New South Wales. Remote access will also be developed for structural instrumentation at the Replacement Research Reactor, due to be operational in 2006, and the Australian Synchrotron which is expected to be operating by 2007.

Remote access may be active or passive and the choice depends on the nature of the instrument and the nature of the experiment or measurement it supports. Passive observational access is appropriate for relatively quick experiments or measurements that do not require or depend on critical assessments based on initial 'screening' measurements, or experiments that are too complex for automation and remote control. Examples include X-ray absorption spectroscopy, small angle X-ray and neutron scattering (SAXS/SANS), X-ray and neutron powder diffraction, reflectometry, secondary ion mass spectrometry (SIMS), scanning Auger microscopy (SAM) and scanning probe microscopy. Active remote access is appropriate for relatively lengthy measurements where early decision making may critically determine the outcome of the experiment, and is particularly beneficial where the experiment is amenable to automation. Single crystal diffraction (X-ray and neutron) is highly amenable to automation and a number of systems are now in operation around the world, including commercial systems. The technique also requires early assessment of the sample quality before proceeding with an experiment. Similarly, micro-probe and tomography experiments are suitable for active remote access, and some may be amenable to automation.

Active remote access will require the development of a user management system to control access; user input to the measurement process; experiment tracking, logging and accounting; data delivery; and data storage. In a generic system each instrument would have a 'proxy' server that communicates with the instrument using protocols required for that instrument. User control for each instrument will be site specific. A workstation would mediate authentication and communication between the client computer (PC, MAC or other) and the instrument server. The workstation would also control the distribution of compressed data, including delivery of instrument and webcam video to the client computer as compressed streaming JPEG. The user interface will likely be written in Java for web browser delivery, and will be designed to be as portable and extensible as possible.

The MMSN has participants with internationally outstanding expertise to help advance its remote instrument access program. Funding for the equipment required to enable remote access will be sought through concerted Linkage Infrastructure Equipment and Facilities (LIEF) funding applications. Equipment will include sample mounting robots, data storage systems, web camera systems (eg. www.axis.com) and internet communication devices such as PolyCom systems (www.polycom.com). Remote access infrastructure will be acquired and installed progressively, culminating with installations at the major facilities. Grid resources for the Replacement Research Reactor and the Australian Synchrotron will be pursued separately and incrementally, with a view to unification on maturity.

Databases and Visualisation
The intent of the program is to incrementally establish a national molecular and materials structure database service, primarily serving the community represented by the MMSN. In addition to conventional databases the program will explore a Grid-based structure database system, and a complementary Grid based spectroscopic database. The database service will be collaborative in character, with the principal vehicle for collaborative interaction being the world's first Grid-based, mutually interactive molecular visualisation and analysis system. The system will provide synchronised displays to multiple monitors that may be located anywhere in the country, or overseas.

The MMSN database service will be piloted by providing access to the principal databases used by the molecular and materials structure sciences; the Cambridge Structure Database (CSD), the Inorganic Crystal Structure Database (ICSD), the Protein Data Bank (PDB), the Metals Data File (MDF) and the Crystal Data Identification File (CDIF). The national molecular and materials structure database service will be established in close collaboration with the UK's Chemical Database Service (CDS), which provides a comprehensive structure and properties database service at no cost to subscribing academic institutions (http://cds.dl.ac.uk). The Australian database service will be modelled on that provided by the CDS, a task that will be accelerated through visits to Australia of CDS personnel. The performance measures used to monitor the CDS performance will be used to assess the Australian database service.

The MMSN will develop the world's first Grid-based collaborative molecular visualisation system, such that multiple users in differerent locations can simultaneously interact with a synchronised molecular display. The combination of a Grid computing engine for photo-realistic rendering calculations and a thin display client on the user's machine will ensure very high speed and high quality rendering that will maximise the efficiency of collaborative and educational interactions. Multiple network users will be able to choose between all users interacting with the same display on their geographically disparate monitors, or each having multiple display windows on their (large) monitors with each window display being driven by one of the participants. In such a system the same molecule might be viewed in each window (in multiple window mode) or different molecules may be displayed for comparison purposes. The system will provide geometrical analysis capabilities, and multiple rendering options would be provided, including innovative rendering techniques such as the use of Hirschfeld surfaces being developed by Prof. Mark Spackman for visualising molecular interactions. Users would communicate visually and verbally through the network, or with conventional telephone or video conferencing equipment (http://www.polycom.com/home/).

National Benefit

There is international recognition by Governments that future economies will be driven by smart information use, smart devices, smart material use, nanotechnology and biotechnology. There can be no doubt that the molecular and materials structure sciences will underpin these futures. The MMSN will contribute significantly to smart Australian science, and thereby help maximise the returns from a knowledge-based economy.

The MMSN infrastructure program will efficiently accelerate progress in key scientific endeavours, including those identified by the Australian Government (http://www.dest.gov.au/priorities/) as National Research Priority 3: Frontier Technologies for Building and Transforming Australian Industries. The Research Priority 3 specification emphasises the importance of breakthrough science and smart information use, and recognises Australia's strength in fundamental science and key technologies such as biotechnology, advanced materials, and information and communications technology (ICT). The latter is seen as a critical enabling technology in Research Priority 3, which highlights a need to invest in smart information use and data management in contributing to productivity, growth, competitiveness and well being. The goals of the MMSN have resonance in Priority Goal 4 of Research Priority 3; smart information use. The MMSN encapsulates smart and efficient information capture, its distribution, storage and analysis, and resulting knowledge generation.

Knowledge of the three dimensional arrangement of atoms, and the chemical identity of those atoms, in molecules and solid state materials is crucial to understanding their behaviour and harnessing their potential in real-world applications. Such knowledge can deliver profound scientific, technological, social and economic rewards, as recognised in Priority Goals 1 (Breakthrough Science), 2 (Frontier Technologies) and 3 (Advanced Materials) of Research Priority 3. Understanding the biological process that support and shape life and the disease and degenerative process that may threaten life, depends increasingly on a detailed knowledge of bio-molecular structure of proteins, viruses, DNA complexes and enzymes (eg. the SARS protease). Rational medicinal drug design, which produced the first anti-flu drug Relenza (developed by two MMSN participants) and the HIV protease inhibitors, would be impossible without a knowledge of molecular structure. The determination and rationalisation of the relatively small atomic structures comprising micro-magnets, microporous and mesoporous materials, hydrogen storage materials, novel metal oxides, ceramics, superconductors, minerals, 'smart' materials, piezoelectric materials, magneto- and electro-rheostatic materials, photonic devices, information storage devices, molecular switches and sensors, biomimetic materials, and pharmaceutical materials is crucial to their development and utilisation.

The network embraces Priority Goal 5 in using breakthrough technology and the internet to host a collaborative research environment in which Australian distances are no longer of consequence or disadvantage. Building a national collaboratory will promote an innovative scientific culture, creating "structures and processes for encouraging and managing innovation". Network meetings, both real and virtual will stimulate and add vibrancy to the research culture.

The MMSN is broad in scope and seeks to capitalise on still emerging science and technology. It is unquestionably risk taking, but, without doubt, will significantly impact research central to the promise of a smart economy. Further, the bringing together of research teams within the MMSN will catalyse dynamic new collaborations, breaking down existing barriers and opening unexpected frontiers in research.

The MMSN is supported by ARC Research Networks Seed Funding.

ARC -- Australian Research Council