How does it change the research?

   Rubin: Once you have a genome sequence, for instance, if you have any piece of DNA, you can determine a sequence of 25 or 30 nucleotides and immediately know where it comes from and what its complete information content is. It makes lot of experiments that were extremely time-consuming in the past go very rapidly. Also, a lot of times in biology you have complexes of interacting proteins. Suppose you know one protein and you make an antibody against it and use it as a probe to go in and isolate that protein in its normal context in a cell. But now you want to discover what other proteins are present in the complex with it. This can be a very daunting task. If, however, you have the sequence of the whole genome, you can take all those other proteins, treat them with a protease, take the little fragments of proteins that result, put them in a mass spectrometer, and determine their molecular weights. Normally that’s not enough to identify the proteins without other information. But if you have the whole sequence of the genome—6,000 genes in yeast—you can directly predict the amino acid sequence of the proteins that are encoded from the DNA sequence. And you can use a computer program to predict all the peptides you would get, along with their molecular weights, and then compare them with what you have. It narrows down the possibilities tremendously. It’s now a simple task to look up in a table from one to another and figure out what proteins are in your complex.
   These types of techniques have changed the way people do science using organisms with sequenced genomes. There’s something nice about knowing all the components you’re working with. If you have a mutation in a gene and don’t get exactly the phenotype that you expected, you can ask if there are other genes in the genome that are very similar to that and which might have a related function. It’s like having a periodic table of the elements. You know going in what all the elements are in the table. It enables a whole range of technologies. So when people saw what was going on in yeast, they realized they wanted a complete Drosophila genome. The same people who were complaining about how we were wasting money sequencing the genome were now complaining about why it wouldn’t be done in one year instead of three.

Originally, as you've just said, estimates projected that it would take about three years to complete the genome. Now you’ve made a deal with Craig Venter’s new company, Celera Genomics, to sequence the genome faster. How did that come about?

   Rubin: In May of 1998 we were at a meeting at Cold Spring Harbor. He came to me and said he would like to test his whole-genome shotgun approach on something smaller than the human genome. No one knows whether that will work. We know it works on a genome the size of bacteria, which is roughly one-twentieth the size of Drosophila’s. So he thought that sequencing Drosophila would provide an intermediate test of the technique. This was just after he set up Celera. There were a number of issues that we had to resolve, because this would be a collaboration between a private company and government-funded effort that is supposed to produce data for the public domain. But now they’ve actually started to sequence Drosophila, and if everything goes well, our combined efforts could finish the genome. Although like any scientific experiment, the end result is somewhat unpredictable.

How do you see all these sequenced genomes changing the nature of biological research in general?

   Rubin: Basically, what we were doing with all developmental genetics was finding out the functions of genes. When we have the whole genome, first of all, a lot of things become more efficient. Going from a mutation identified in a genetic screen to actually isolating the gene is much much faster. The other thing is that there are 5,000 people worldwide working on Drosophila, and Drosophila probably has less than 15,000 genes. So that’s one human being for every three genes. If you give those people very efficient tools for figuring out the functions of genes, you can do it in a massively parallel way.
   The other way that genome projects change things is this: inside of cells are a very large number of complicated mechanisms, each much more like a network than a pathway. To really understand those inner workings we are going to have to look at many genes at once. That’s not the way the human brain likes to work. So we’re going to need computers and more model-building ways of analyzing this data rather than just one person with his or her unassisted brain trying to figure things out. That is where the genome projects really come into play. They enable us to know all the genes so we can look at all of them at once and see what they’re doing.

Is this where micro arrays will come in handy?

   Rubin: Yes. They’re really just miniature versions of things we have been doing in biology for a long time, but with much higher throughput. You can start with the RNAs expressed in a cell at a given time and, by annealing them to your immobilized array of genes, you can determine what level of expression you have for each of the genes in the array. So you can rapidly build up knowledge about how genes are regulated during development. Now imagine that you do this under 100 different conditions. If you have 10,000 genes under 100 different conditions, you have a million data points. But the human brain doesn’t look at a million data points and make models. You need computers and clustering algorithms and the like. Nowadays, with the ways of acquiring data that the genome project has enabled, people are generating data much faster than they can think about it. It’s exactly the opposite of the way it used to be. Five years ago you spent six months doing some experiments and then you could probably interpret them yourself over a weekend and decide what to do in the next six months. Now it’s the reverse: you can do the experiment over a weekend that generates data that you can think about for the next six months. It’s a paradigm shift in the way at least a certain part of biology works.

It sounds like you’re predicting the existence of a new field of theoretical biology?