Current projects

• Fingerprinting for crystal structure clustering
• Quantitative HIV-1 particle tracking in target cells
• STM4 molecular visualization toolkit
• CSCS users visualization support
• Teaching and knowledge sharing

Exploratory projects

• The perfect tool project
• Representation in chemistry
• Integrated discovery cycle support
• HCI issues in visualization tools

Current projects

Fingerprinting for crystal structure clustering

USPEX is a method developed jointly by Artem R. Oganov and Colin W. Glass and implemented in the same-name code written by Colin W. Glass and Andriy Lyakhov that enables crystal structure prediction at arbitrary P-T conditions, given just the chemical composition of the material.

Each USPEX run produces thousands of structures, but in practice many of them are the same structure, perhaps described using a different unit cell or a different spatial orientation of the cell. Currently the extraction of unique structures from the results of an USPEX run is an intensive manual labor, consisting mainly in judging equality from side-by-side visualization of pairs of structures.

The fingerprinting project aims at the creation of an automatic structure comparison and clustering method targeted to this specific usage and not necessarily intended to be a solution to the general problem of finding equivalent structures for generic molecules.

The project name derives from the first step of the algorithm: here a descriptor that captures the true spatial configuration of the atoms, i.e. a fingerprint, is associated to each structure. Then on the basis of an abstract distance between those fingerprints the structures are clustered in groups of approximately equal structures.

We have developed an application inside the molecular visualization toolkit STM4 that clusters structures produced by a USPEX run and provide various diagnostic instruments to experiment with various algorithms.

The future phase of the project will be the inclusion of the structure clustering mechanism inside USPEX to improve its discovery capabilities.

• Project pages
• Prof. Oganov group pages
• USPEX

Quantitative HIV-1 particle tracking in target cells

To investigate the mechanism used by a viral particle after it enters a cell to move to the cell nucleus, the group produces time-lapsed 3D images of the infected cell after attaching a fluorescent marker to the viral material. A subsequent image analysis extracts the 3D trajectories of the fast-moving viral particles and then statistically analyzes them to infer the mechanisms used to reach the cell nucleus.

This project aims to develop image management and analysis tools in partnership with CSCS to allow real time image acquisition and analysis to be done automatically and with minimal user input.

This is key in order to be able to move the field of cellular virology from an extremely time consuming, slow field to a high through put field of studies that can be effectively exploited for drug development and patient care.

Another project goal is to select and validate a data management and integration environment that could be used also in other fields including studying other stages of the viral life cycle, other viruses and physiological cellular processes such as vesicle trafficking.

Currently we are evaluating and using OMERO, a multi-site collaborative effort among academic labs and a number of commercial entities that produces open tools to support data management for biological light microscopy. Designed to interact with existing commercial software, all OMERO formats and software is free, and all OMERO source code is available under GNU public “copyleft” licenses.

• Research group pages
• Open Microscopy Environment (OME)

STM4 molecular visualization toolkit

STM4 is a molecular visualization toolkit built on top of AVS/Express. Its main goal is not to replace existing tools, but to provide a platform on which new, advanced or unusual visualization techniques can be implemented.

The main STM4 strengths are its quick prototyping capability, which permits a tight interaction with the users to elicit their real needs, and the ability to use standard visualization techniques together with the more usual chemistry visualization ones. Those capabilities make STM4 a tool that does not overlap with other visualization tools, like Molekel, that fit more the visualization needs of an end user.

STM4: the evolving molecular visualization toolkit

This year the molecular visualization toolkit has undergone a major restructuring to better support new user requests for new and advanced visualization techniques. The new version has at its core an expanded Data Model to better describe molecular structures and a new set of visualization modules created to support specific user projects.

Besides the STM4 visualization support to chemistry related research users, the toolkit is the base for specific visualization projects letting them concentrate on the development of project-specific functionalities. For example, within STM4 has been implemented a crystal structure classification and analysis tool for Prof. Oganov's USPEX tool.

• STM4 web pages

CSCS users visualization support

I provide visualization support to various CSCS' Large User Projects (LUP) in the chemistry area. This is a rather open ended project focused especially on post-processing, movies and images production.

This project is strongly linked with developments inside the molecular visualization toolkit STM4. In fact the reason for the development of new functionalities inside STM4 derives from needs arising during user visualization support.

Currently I'm supporting a researcher form the Prof. Parrinello research group producing images and movies about “Study of the diffusion of vacancies in solid LiAl alloys”.

Teaching and knowledge sharing

Part of my time is dedicated to the preparation and delivery of courses on "Visualization foundations" and "Visual communication". Refer to my teaching page for a complete list of courses taught.

One of the most enduring collaborations is with the CINECA Summer Schools. This collaboration in turn is generating other opportunities (like the CESIC master in summer 2007) that spread the acknowledgment of CSCS as a visualization excellence center.

Other talks and tutorials were more focused, for example the AVS/Express workshop made for the Swiss AVS distributor.

Exploratory projects

The perfect tool project

Chemistry is a visual science, but chemistry visualization tools lag behind. This project tries to collect from the existing tools ideas and drawbacks to define and propose what a perfect tool should provide.

The name “The Perfect Tool Project” is an ironic name with a double meaning:

1. Looking at the existing visualization tools with a non-chemist eye, seems that they are far from perceptions and worse, they lags behind the chemistry reliance on graphical representations
2. It is intended to be an “igniter” for both visualization experts and chemistry researchers to make both fight the status quo and ask for new developments and functionalities

The project has just started. Currently you can access the first report and the list of tools tested.

The project idea is to define a series of themes related to visualization usage in chemistry and prepare small example applications to collect researchers' feedback. Those feedbacks will be initially collected to define the characteristics a “perfect” chemistry visualization tool should have.

But more that a new tool design, the project what to provide an “excuse” to make the researchers think outside their usual tools about what visualization could provide to them to make their work more effective.

More details are available on the project page.

Representation in chemistry

1. Chemists usually deal with a narrow range of data types.
2. Chemistry visualization uses a limited, traditional set of representations for those data.
3. I am convinced that more complex data types and unusual representations could help bring new insights in the chemistry research.
4. Otherwise the risk is this: “A stagnant set of representations limits the way scientists think about their models and thereby limits potential insights”.

So I collected this material as one of the inputs for a research underway about representation in chemistry and the support or constrain provided by the current chemistry visualization tools.

My hope is that this work could help me in creating and proposing new visualization tools and techniques. And, maybe, catalyze new ideas for the chemists I collaborate with.

The project has created an initial collection of representation methods in chemistry (and the parallel collections of errors in visualization tools) that spots the almost complete uniformity of representational methods between tools.

Integrated discovery cycle support

Is the scientist who makes the discoveries, visualization is only an internal interface in the discovery cycle.

But around this cycle, there are other activities that are critical for the discovery outcome, but still neglected by the current tools like:

• Idea generation and recording
• Interaction with computational facilities
• Consulting with peers
• Access to literature and domain knowledge
• Results dissemination

Also the simple visualization process as usually modeled, see the Robertson model for example, differs markedly from the real life usage of a visualization tool.

The project starts collect ideas from other fields, like games, HCI and creativity support with the goal of providing input to the design of CSCS visualization tools.

HCI issues in visualization tools

To prepare the possible future development of our own molecular editor, I started collecting ideas and studying the field of Human Computer Interaction (HCI). Besides theory I, together with Ugo Varetto, am experimenting with possible implementation technologies like wxWidgets for the GUI and Open Scene Graph for the graphical scene.

A subfield worth investigating is the use of ideas from the videogames field. For this a growing bibliography starts accumulating.