Sunday, May 30, 2010

Artificial News

I have been meaning to write something puncturing the hype about "artificial life", but I see that Craig Venter already wrote a clear explanation of what his team actually did for the Wall Street Journal:

In 1995, we reported the DNA sequences for the first two cellular genomes. Nowadays genome sequences, which contain the genetic instructions for an organism, are routinely obtained and deposited in computer databases.

Last week, we reported that this process can be reversed. The digitized DNA information of Mycoplasma mycoides, a simple bacterium, can now be brought to life.

Notice the passive voice of that last sentence. Venter and company did not, themselves, bring the DNA "to life", they put it in a fully functioning cell that did that for them.

So what is new and unique about what we did? The process of synthesizing a cell began at a computer. We started with the more than one million letters of genetic instructions for Mycoplasma mycoides, and then made slight modifications to its DNA sequence. First, we deleted 4,000 letters, which removed the function of two genes. We then replaced 10 genes with four "watermark" sequences. These watermark sequences are each over 1,000 letters in length and can be decoded to reveal the names of people, famous quotations and a website address. The entire sequence of DNA letters was then partitioned into 1,100 pieces, and each was synthesized using four different bottles of chemicals that make up DNA. These DNA fragments were designed such that adjacent pieces contained an 80-letter overlap, which facilitated the assembly process by providing unique regions where the synthetic pieces could join.

The synthetic Mycoplasma mycoides genome was assembled by adding the overlapping DNA fragments to yeast. Once inside a yeast cell, the yeast machinery recognized that two DNA fragments had the same sequence and assembled them at this overlapping region. The genome was not assembled from all 1,100 pieces at once but rather in three stages: 1,000 letters to 10,000 letters, 10,000 letters to 100,000 letters, and finally 100,000 letters to complete the 1.08 million letter genome. This assembled genome is the largest chemically defined structure ever synthesized in the laboratory.

. . . nor did we create life from scratch. We transformed existing life into new life. We also did not design and build a new chromosome from nothing. Rather, using only digitized information, we synthesized a modified version of the naturally occurring Mycoplasma mycoides genome. The result is not an "artificial" life form. It is a very real, self-replicating cell that most microbiologists would be unable to readily distinguish from the naturally occurring counterpart without the aid of DNA sequencing.

So what they did was build, with the help of yeast cells, a very big molecule. Because that molecule is DNA, people have gotten all excited about "making life," but DNA is not life. It is just one molecule, just one part of the millions of parts that make up even the simplest cells. A very important part, but by itself DNA is not life. Without a fully-functioning cell to carry out the instructions it encodes, it just lies around for a while and then decomposes.

I also pass along this piece of satire, about "journalists create world's first artificial news story."

No comments: