The physician and writer Lewis Thomas claimed in his book of essays "Lives of a Cell" that, if the mystery of biological development was ever solved, he would hire a fleet of skywriting planes to carve giant exclamation points in the sky until his money ran out. Luckily for Thomas, perhaps, the mystery remains unsolved, but progress has been made since 1974 in elucidating exactly how an organism's dividing cells begin to be partitioned into ear cells and stomach cells and heart cells. You may remember from high-school biology that your genome consists of coiled DNA that acts as a recipe of sorts (see, even I'm doing it!) for different proteins. First, DNA is transcribed into a single-strand copy of RNA. We now know that this RNA contains, mysteriously, segments called "introns" that do not contain useful information. These segments are snipped out and thrown away and the remaining segments, called "extrons", are joined together to form mRNA. This mRNA is then translated into a protein by the cell.

Different genes are "activated", that is, will be expressed and create proteins, depending on the chemical milieu in which a given cell is immersed. Moreover, certain genes called "homeobox genes" are able to switch on or off sets of other genes. For example, one particularly prominent set of homeobox genes, called the Hox genes, determine the position of various segments of the body in a developing fetus or larva.

So far I have assumed that each gene encodes for a simple protein, the classical view of how genes are expressed. It turns out, however, that many genes can actually be translated into a number of different proteins. (The information in the remainder of this post comes from the article "The Alternative Genome", by Gil Ast, in April's Scientific American. It's well worth reading in full.) This phenomenon was first discovered in the 1980's by Randolph Wall at the University of California at Los Angeles and Tom Maniatis and colleagues at Harvard Unviersity, who discovered that the genetic machinery can, if it desires, include introns in the mRNA and exclude entrons from the mRNA. This effect, called "alternative splicing", allows a single protein to encode any of a number of different proteins. Originally, alternative splicing was thought to be relatively uncommon, but recent research has suggested it takes place much more often, especially in more complex organisms: as many as three quarters of human genes may undergo alternative splicing, and on average each of our genes can be spliced into mRNA in about three different ways.

The completed sequencing of the human genome has revealed that humans have only 25,000 genes--15,000 less than corn, a relatively humble organism. This is, to be sure, a small cookbook, and not much of a blueprint. However, if you view the genome as a computational device rather than a list of proteins, this tiny number begins to make more sense--the human genome has simply learned to more efficiently store information by allowing each gene to encode a number of different proteins. All of these recent revelations point towards a need for a new metaphor or set of metaphors for the Cookbook of Life, which is an exercise that I'm leaving for the reader.