The "Google" Collection of Chemical Reactions

From our friends at Northwestern University in Evanston, Illinois:  

Northwestern University scientists have connected 250 years of organic chemical knowledge into one giant computer network -- a chemical Google on steroids. (Called Chematica, t)his “immortal chemist” will never retire and take away its knowledge but instead will continue to learn, grow and share.

 A decade in the making, the software optimizes syntheses of drug molecules and other important compounds, combines long (and expensive) syntheses of compounds into shorter and more economical routes and identifies suspicious chemical recipes that could lead to chemical weapons.

The Chematica network is unique in that it connects approximately 7 million chemicals with a like number  of chemical reactions.  Essentially, they have linked all known chemical compounds with the reactions between them.  It is a monumental repository of chemical reactions/methods.

Another added value of the project is the development and use of algorithms to search and analyze the data.  This is a very useful capability for the research community.

Details of the system are published in the August 6 issue of Angewandte Chemie.

Chematica is being commercialized, which speaks to the relative maturity of the research and their trust in its capability.

While they are not characterizing Chematica as a model or a simulation, the ability to reference this data may one day be useful to modelers seeking to accurately represent physiological reactions within the human body.  Once again, it's all about the data.

Digital Health

Why does a blog on medical modeling and simulation collaboration care about the various ways to track one's health via commonly available technology?  Why do we care about electronic health records and other historical information relative to one's health?

It's the data.

Models and simulations are data-hungry tools.  Data is sustenance to models and simulations -- without it, they have nothing to feed the algorithms and queries that make up most models and simulations. For those of us still wrapping our heads around M&S as a tool, think of data this way:

Algebraically, the polynomial X+Y=Z is not a terribly difficult problem -- assuming you have at least two numbers to substitute for X, Y, or Z.  Without at least two numbers - the data - this problem becomes unsolvable.

One of the great but mostly unspoken benefits of electronic health records is the potential to gather and use data from individuals, and in the aggregate, communities that physicians and health professionals can leverage in the increasingly complicated task of diagnosing and treating people accurately and with economy.  Electronic health records, as they exist today, are incomplete and not interoperable.  They certainly don't collect information such as "I felt like crud today" or "I ran six miles today and my left knee hurts and I'm running on 18 month old neutral last running shoes and ....." all which may be useful to a diagnosis.  Most people don't track themselves completely - I'm of the opinion that our personal health memories are mostly founded in personally catastrophic events.

Capturing personal data may appear fairly narcissistic or obsessive-compulsive, but it might play an important role in predicting potential health issues or elaborating upon historical indicators that contribute to issues.  As technology progresses, there is potential for much of the required narcissistic behavior to be replaced by an autonomous collection of much of this data.  I suppose we'll figure out whether we like that or not, down the road.

Photo by Karen Roe/flickr/CC

Wired recently published an interesting article out using technology to track your health.  The article presents a basic blueprint to tracking the things that may help you figure out why you're gaining weight or losing weight or tired or whatever.  Their article also outlines a number of websites that can assist with keeping track of yourself, health-wise.

Something I'm in the process of evaluating is HealthTap.  HealthTap is an online service that matches you with physicians in your area, as well as answers to personal health questions.  According to the site, "HealthTap is for people who want instant access to trusted health information from top doctors."  The site provides you with the opportunity to input a vast quantity of information about your health history as well as input daily information.

While Google Health tanked, the idea was consistent with many of the online or digital oriented health tracking businesses out there.  I would guess that Google will get back into this game further down the road.

What's the benefit of any of these sites, today?  The privilege of being an "early adopter" and digital access to your personal information.  At the NCCMMS, we see these as future assets that, with your permission, we can leverage to verify and validate the performance of analytic and predictive models that   will be used to educate as well as assist with diagnosis and treatment.

Healthcare Analytics

According to Wikipedia, Analytics is "the discovery and communication of meaningful patterns in data. Especially valuable in areas rich with recorded information, analytics rely on the simultaneous application of statistics, computer programming and operations research to quantify performance."

Medical modeling and simulation, as an analytic tool or capability, are an integral part of healthcare analytics.  Certainly, data analysis can be performed without models; however, models provide the flexibility to evaluate multiple threads, look for optimizing characteristics, or chase theories, for example.

In my mind, healthcare analytics is a natural extension of medical data analysis enabled by modeling and simulation.

The Promise of Healthcare Analytics article.

Free Report Download

NCCMMS in the News

Published on HamptonRoads.com on 12 August, 2012| PilotOnline.com (http://hamptonroads.com)

Simulating the Future for VMASC

Proponents of modeling and simulation all seem to agree that the possibilities within the industry - for jobs, for improving quality of life, for solving complex problems - are virtually limitless.

The puzzle, so far, has been how to realize that potential by actually drawing people into the discipline and getting businesses to see the value in it.

Members of the governor's advisory council on modeling and simulation reiterated the scope of the challenge last month, and the dearth of American workers with skills for such technical jobs.
But, they noted, efforts are under way to address it directly. Not surprisingly, many of those efforts are coming out of Hampton Roads, the commonwealth's hub for modeling and simulation. Officials are trying to blunt the 2011 closure of Joint Forces Command and impending defense-related cuts by redirecting modeling work away from the military, and emphasizing its potential.

John Sokolowski, executive director of the Virginia Modeling, Analysis and Simulation Center at Old Dominion University, said researchers there created an iPad app, the "M&S Cube," and released it for free through iTunes three weeks ago to broaden awareness of jobs and resources available to students and professionals. The industries range from transportation to agriculture, construction to manufacturing.
Meanwhile, VMASC is working with CBS affiliate WTKR to air TV commercials statewide touting careers in modeling and simulation. In late September, VMASC is set to hold an open house and host a four-day industry conference.

By late fall, Chmura Economics and Analytics is scheduled to release a study identifying industries in Hampton Roads that are prime targets for expanding modeling and simulation.

One of the most obvious is health care, where local researchers are working on projects that could draw national attention by the end of the year. While Eastern Virginia Medical School is looking at whether to merge with the College of William and Mary, its researchers have spent years working with ODU and VMASC on tech-related projects.

The latest is the National Center for Collaboration in Medical Modeling and Simulation. Long title notwithstanding, the center could reshape the way health care providers, manufacturers and medical centers and schools evaluate and shop for technology.

One way, according to Sokolowski, is through a system that serves as a kind of Consumer Reports-style clearinghouse for reviewing simulation equipment used in hospitals and training centers.

Researchers might, for example, identify limitations in mannequins and accompanying software used to simulate surgery patients, whether the technology did what it purports to do and whether students learned more efficiently through one device than another. Findings would be used to inform hospital and medical school administrators and drive competition. Researchers hope to draw support from major medical associations by November or December, he said.

Another effort mirrors the Media Lab at the Massachusetts Institute of Technology. The project, dubbed "The SimLab," will link academic and industry experts in development of new devices, programs and applications. Officials hope to pull private investors on board within a year, Sokolowski said.

Geoffrey Miller, director of the EVMS Center for Simulation, said researchers there are working toward creating "refresher" apps for physicians, and designers and engineers are developing an app that shows general safe patient-handling tips, such as proper lifting technique.

In the abstract, that app may not sound very appealing to businesses. But the underlying idea applies to anyone, in any industry, who lifts heavy things or works in a place where there is risk of injury. And if development and testing can raise awareness and help reduce workplace injuries, business owners would recognize it as a way to reduce injury claims and save money.

For years, those kinds of possibilities haven't drawn the attention that they should. If VMASC has its way, that is about to end.

Shawn Day is an editorial writer for The Virginian-Pilot. Email: shawn.day@pilotonline.com.

---

Source URL (retrieved on 08/13/2012 - 09:58): http://hamptonroads.com/2012/08/simulating-future-vmasc