Craig Venter on How the Genomic Era Is Just Starting

2000 The human genome is decoded for the first time.

There have been lots of stories written about all the hype over getting the genome done and the letdown of not discovering lots of cures right after. The biggest finding when we announced the genome in 2000 was that we only had 20,000-some-odd genes. Some people thought that because we were humans, we had to have more than every other animal. We had to have 10 times more, 100 times more. Everybody wanted there to be this nice, linear path where you have a gene for each trait—you know, there’s a gene that codes for your nose. It’s so naive.

Only now are we really starting to appreciate all the changes that have occurred over the past 15 years. On the computing side and the sequencing side, both of those just passed a cost and performance threshold that will allow for a major impact on medicine. We’re just now where I wanted to be in 2000. In the last three years, we’ve seen some impact on medicine with the discovery of the driver mutations in cancer. If you have lung cancer, the most important thing you can know is your genetic code. Roughly 4 percent of people with lung cancer will have this ALK gene translocation that seems to indicate that Pfizer’s drug has a better-than-60-percent chance of shrinking your tumor. Instead of a blockbuster drug that you can give to everybody that has cancer, you end up with a drug that you give to 4 percent of the people we know can benefit from it. That is a fundamental change in medicine.

There’s been a huge increase in the last three years of people measuring these driver mutations, but revolutions in medicine are slow. My understanding is only about 3 percent of cancer patients in the U.S. get genetic screening of their tumors. We have the tools to do it now, but the physicians don’t have the training and the know-how to help get this implemented to benefit their patients. The only way to come up with all of the cures that were promised is to sequence large numbers of genomes. For each gene in your genome, you quite often get a different version of that gene from your father and a different version from your mother. We need to study these relationships across a very large number of people. It’s going to be important to know what the variant is you got from your mother and from your father, and whether that correlates with 30 other variants across the genome that are associated with susceptibility for a certain type of cancer, for example.

That’s why we started my new company Human Longevity. We’re trying to sequence 40,000 genomes in the next six months and then scale up to 100,000 a year. We have a goal of getting to a million genomes by 2020. It’s around $1,000 to $1,500 to do a genome today, but we’re counting on continued Moore’s Law-type change to take it down to a few dollars per genome.

We’ve proven that DNA is our software, the software of cells. But genomics can’t really, truly impact medicine—get us to preventative medicine, get us to new treatments—until we can truly read that software. We have less than 1 percent of the information that people will have in the future. So to me, we’re just starting the genomic era now.—As told to Ashlee Vance

    Before it's here, it's on the Bloomberg Terminal.