Computational creativity and the future of AI

'Robotic biologist' crunches raw data and formulates equations


October 17, 2011

A microformulator, designed to allow ABE to perform experiments without human intervention...

A microformulator, designed to allow ABE to perform experiments without human intervention (Photo: Wikswo Lab, Vanderbilt University)

While some people may have been impressed (or intimidated) by the recent development of a system that automatically raises and analyzes cell cultures, it turns out that another facet of the biological research process may also be going to the machines. An interdisciplinary team of researchers recently demonstrated a computer system that is able to take in raw scientific data from a biological system, and output mathematical equations describing how that system operates - it is reportedly one of the most complex scientific modeling problems that a computer has solved entirely from scratch. While the system is known affectionately as "ABE," it is also being referred to as a robotic biologist.

ABE, or the Automated Biology Explorer, was developed through a collaboration between scientists at Vanderbilt University, Cornell University and CFD Research Corporation, Inc. It's a modified version of an existing piece of software known as Eureqa, which was created at Cornell and released two years ago. Eureqa was originally intended for use in the designing of robots, where it would eliminate the trial and error process. In an early demonstration of the system, it was able to independently identify the basic laws of motion, by analyzing the movements of a double pendulum.

When Vanderbilt's Prof. John P. Wikswo heard about this accomplishment, he realized that the system could probably also be used to solve biological puzzles. He contacted one of Eureqa's creators, Cornell's Hod Lipson, and the collaboration that ensued resulted in ABE.

In a test of their new system, the scientists fed ABE a data set that corresponded to the raw measurements a scientist would make when observing glycolysis, the primary process that produces energy in a living cell. To keep ABE on its toes, they even introduced an element of random error to ten percent of that data. Even with that handicap, however, ABE was reportedly able to arrive at a set of equations that were nearly identical to those already known to be correct.

The researchers are now developing lab-on-a-chip technology that will allow ABE to conduct experiments on its own. They are starting with a microfluidics device for testing cell metabolism.

Another "robot scientist" capable of running its own experiments already exists at the University of Wales at Aberystwyth, although unlike ABE, it is reportedly not capable of starting completely from scratch.

A paper on ABE was recently published in the journal Physical Biology.

About the Author
Ben Coxworth An experienced freelance writer, videographer and television producer, Ben's interest in all forms of innovation is particularly fanatical when it comes to human-powered transportation, film-making gear, environmentally-friendly technologies and anything that's designed to go underwater. He lives in Edmonton, Alberta, where he spends a lot of time going over the handlebars of his mountain bike, hanging out in off-leash parks, and wishing the Pacific Ocean wasn't so far away.   All articles by Ben Coxworth
1 Comment

The singularity approaches.

Joel Detrow
17th October, 2011 @ 04:58 pm PDT
Post a Comment

Login with your gizmag account:

Or Login with Facebook:

Related Articles
Looking for something? Search our 31,306 articles
Recent popular articles in Science
Product Comparisons