Glasshouse injects 3D representation of data into a virtual world
Glasshouse by Green Phosphor is a gateway which can take a database query or a spreadsheet and place a 3D representation of it into a virtual world. Users can see data, and drill into it; re-sort it; explore it interactively – all from within a virtual world. Glasshouse produces graphs which are avatars of the data itself.
Virtual worlds are environments where avatars can observe and interact with eachother. This is a wonderful advance for human to human interaction; and it is also a wonderful advance for human to program interaction. Imagine a useful program residing somewhere on the Internet… and to interact with it, all you need to do is invite its avatar into a virtual world with you. The avatar of the program is its 3D user interface.
Imagine the program is architectural software and it can generate into the environment anything you are designing. Suppose you have it put a house into the virtual space, and all the components of the house are interactive – you can click on a door and change the type of door; you can click on the floor and choose a new kind of tile from an actual inventory database out there on the net. Imagine interacting with a molecular model of a protein and when you click on various amino acids and ligands they present menu options for modifying the model or cross-referencing with others.
Imagine a program which searches the web for you. The program’s avatar looks like a librarian and follows you around, and when you say “find me the most authoritative website on bodybuilding” it injects a web page into the virtual environment. Imagine interacting with software as you would with a personal assistant.
All this is possible through the use of protocols such as CICP and MXP, which allow external entities to place interactive content into virtual worlds. All the logic and machinery stays outside the world and just the user interfaces (objects) go inside the world for you to interact with. This is the vision that led me to create CICP – and the particular way Green Phosphor is using CICP is to create interactive representations of data.
Green Phosphor’s product is called Glasshouse, and it is a gateway which can take a database query or a spreadsheet and place a 3D representation of it into a virtual world. Users can see data, and drill into it; re-sort it; explore it interactively – all from within a virtual world. Glasshouse produces graphs which are avatars of the data itself. We’ve tailored the system for the use of biotech companies, specifically for drug discovery and development. Dr. David Resuehr, a molecular biologist, recently joined Green Phosphor as our Chief Scientist. We’ve just now entered our marketing phase – for more information please visit http://greenphosphor.com/?location=Biotech.
The power of interacting with data in 3D has to be experienced to be believed. Some day soon many people will have jobs which require SQL skills and virtual world skills; these ”data wizards” will turn information into intelligence within the 3D environment. They’ll facilitate a data-driven decision-making process which will hopefully reduce some of the short-sightedness and stupidity we’ve seen lately in many of our critical institutions.
Jani: What do you refer to when you say: “…will hopefully reduce some of the shortsightedness and stupidity we’ve seen lately in many of our critical institutions.”? Could you give an example?
Ben: I have a couple examples which I’m sure won’t offend anyone and a third which may. First off I believe that visualization of money in/money out could have turned authorities on to the fishy accounting Enron was up to, and caught them earlier. Perhaps better visualization would have revealed Madoff’s ponzi scheme as well. It’s a matter of seeing the big picture; money must come from somewhere and be tied to intrinsic value at some point. My third example is similar, but rather than money in/money out it is about energy in/energy out. I believe proper visualization of the process for producing ethanol fuel, including energy cost of fertilizer, would reveal the uselessness of subsidizing cornbased ethanol as an alternative to fossil fuels. Much of the value of looking at data has to do with getting in touch with history rather than trying to look at realtime feeds with everfiner granularity. There is tremendous value in looking at the past in objective ways – and the best objectivity comes from hard, unadulterated data. Combine data visualization and visible process simulation with massive collaboration – crowdsourcing, masterminds formed by teams – and the human race can be more intelligent.
Jani: I’d like to get a quote from Dr. David Resuehr.
DrDavid: For science, the 3-D environment may really have a very prosperous future. We know that the discovery of many drugs was more or less coincidental like for instance when Fleming discovered Penicillin. Others were specifically and tediously designed, tested and created. Insufficient testing can – as it happened with Thalidomide – have terrible consequences. During the 50’s and early 60’s thousands of children in Africa and Europe were born with severe malformations because their mothers had been given this drug against morning sickness.
What does all this have to do with virtual worlds and 3D environments? The connection lies in the fact that molecules are not flat but have a (sometimes very elaborate) three dimensional structure. Here is the anecdote: Thalidomide exists in two isomers (enantiomers), which are mirror images of each other, like your right and left hand. The problem was (as was found out later) that one of the versions, the ‘S’ enantiomer, was teratogenic [causes malformations]; but the ‘R’ enantiomer was an effective sedative. I like to refer to the ‘S’ enantiomer as the ”evil twin”. Had the developing scientists been able to display and most of all discuss this and its possible implications with their peers, maybe they would have been more diligent and could have prevented the release of this cocktail and its devastating effects for thousands of people. Clearly it is impossible to recognize and tell apart a “bad” molecule from a good one by just looking at it, but having it dangling right in front of your nose in cyberspace may make you think about it a little more thoroughly and prime discussions about it.
DrDavid: Virtual worlds like Wonderland and OpenSimulator offer never before available possibilities for researchers from all fields to interact with each other and display and scrutinize their data in new and more intuitive ways. This platform also offers a link from the people in the lab to the people at the desks and enables everyone to give their input, from wherever in the world they are – many brains are better than one and sometimes it may take a non-scientist to come up with a great idea; after all it’s about collaboration and working as one strong team.
DrDavid: A long-term goal is, in collaboration with bioinformatics and cell biology specialists, to create a complete virtual cell. If we can accurately display and simulate cellular pathways, we may be able to better understand drugs’ molecular actions, explain adverse effects, predict risks, and discover potential therapeutical targets for the development of new drugs.
Ben: I believe one of the keys to intelligence is the elimination of bias and preconceptions. This elimination opens the door to truth. One of the great features of looking at raw data in a virtual world is you can start without any labels or explanations of what the data is. You simply see the patterns, the outliers… then click in and see what the actual datapoints are, and be prepared to accept what is revealed. I’d like to close by mentioning a pilot project in Second Life which is using the free version of Glasshouse to show the effects of various companies’ customer loyalty efforts on revenue. The data was provided by Leslie Pagel and the graphs are on display at Gronstedt Group’s Train for Success2 region. Here’s a slurl: http://slurl.com/secondlife/Business%201/211/48/33.
Just what we need – more avatars in things. Thanks. Thanks for that.
for a slightly more analytical tool with a Serious statistical foundation, try SAS’s JMP program (albeit it is not free). Graphics one typically sees on a MAC available on a Win/PC. And the other aspects/functions in the program are intuitive without killing weeks learning the program. (i do own the program (version 7, but ver 8 now available) and consider it one of the better software investments i’ve made for quite a while now.
the website is http://www.jmp.com and a 30 day trial is available.
JMP Genomics utilizes three-dimensional graphs for the visualization of scientific data such as microarray data, but the 3D graphs are presented within a 2D framework and simply rotate around within a fixed space. Our graphs are rendered within a space in which the user navigates through the use of an avatar, giving the user the ability to intuitively change perspective as desired, and allowing realtime collaboration. JMP Genomics does not have three-dimensional representations of proteins and protein subcomponents, and their relationships. This is the next level, which we are currently building using our core technology – the 3D space allows complex networks of proteins and their interactions to be modelled in a way which was not possible without the virtual world.
I’d also like to add that we fully support the use of tools such as SAS or other stats and analysis packages to process data. There’s no reason not to use whatever you want to come up with a meaningful dataset, and then use Glasshouse to render your results into a virtual world for everyone to see. The tools can be complimentary.