Tuesday, September 06, 2005

Artificial Intelligence Analyzes DNA

I just started a new job this week, and this whole 8-hours-a-day-stuff is taking quite a lot from me. My posts will probably be a little shorter from now on.

Back on topic:

Artificial Intelligence Analyzes DNA

"We were able to feed the patterns into an artificial intelligence computer program developed in my group," Frey says. "The computer analysis identified thousands of instructions and changed our view of how genes work." For example, their analysis showed that a region of the fourth chromosome which was thought to contain four short genes actually contains a single very long gene, which is now thought to be involved in the assembly of large protein molecules in the nucleus. By better understanding this and other genes, researchers hope to learn how these genes malfunction and cause disease.


It's things like these that make possible a fullblown biotechnology-revolution, which is predicted to be from 2010-2020 (after which we will enter the nanotech-revolution). Things are really speeding up. Just read this next quote:

The program also revealed a startling discovery: there are no new protein-coding genes to be discovered - the genetic instructions that are largely responsible for managing cells, determining everything from eye colour to disease. By analysing the data and inferring the most likely genes based on user-programmed variables, the program matched what research has taken 30 years to discover. "This flies in the face of research that says there are many more protein-coding genes to be discovered," Hughes says. "We've reached a milestone in gene exploration."


That '30 year'-comment is a marker for exponential acceleration. I read a lot of news articles that cover breakthroughs in all sorts of areas. And these last few years, it is very common for scientists to say that, something that used to cost a number of years, now only takes a number of months. The DNA chip, for example, allowed scientists to look at the activity of many genes all at once. This allowed them to discover more in the last 3 months, than they had in the last 3 years.

I'd never seen anything, that previously took decades, happen in one run of an AI software program. That's pretty impressive!

We are now able to develop an understanding of life itself. Next time somebody tells you science won't decrypt the language that our own genes speak in a hundred years, just send them over to my blog:

"I think that genomics research is one of the most compelling areas of science today," Frey says. "Many people I talk to, from my seven-year-old son to university students across multiple disciplines, are excited when they find out that we now have the capability to develop an understanding of one of the most fundamental aspects of life."


This is not the first time that humans are being outperformed by machines in the lab. A robot scientist has already outperformed humans in the lab more than a year and a half ago.

An intelligent robot that could free genomics researchers from routine lab chores has proven as effective as a human scientist. The robot not only performs genetics experiments, it also decides which ones to do, interprets the results and comes up with new hypotheses.

AddThis Social Bookmark Button

No comments: