That is surprising, I wasn't aware of this.
We are involved with the Hutch and Hartwell and the NIH/NCI on a huge parallel project that I think is far better conceived. Interesting and somewhat baffling that Lee Hartwell evidently spearheaded both, all within a matter of 6 months or so.
I can't say I'm excited by what I read here. The genes just ain't where it's at. They are likely to tell you rather little of any real value.
The reason is because genes have no direct biological effect at all.
It is kind of like this. Imagine if you wanted to know all the things that people are writing on their computers, using word processing software. You decide to invest enormous time and effort into this by cracking the code. Eventually, you replicate line for line all the code that went into Microsoft Word.
So now what do you know about what people are actually doing with MS Word? Practically nothing, because knowing the code behind software gives you little insight into how that code is being used to produce product with the aid of that software.
Similarly, knowing the sequence of the genes, and the variations among genomes tells you enormous amounts of things, but surprisingly little of it is actually relevant to biology or pathology for a number of reasons. First and foremost, life is a manifestation not of genetic interactions, but of protein interactions. Sure, the genes code for the proteins, but only in a very rudimentary way. Let me give you an example, and this will dramatically demonstrate that biologic (and pathologic) complexity and dynamics have precious little to do with genetic sequences.
There is a gene called DSCAM, discovered here by a friend of mine, incidentally. My friend discovered this gene initially in flies, but it exists in humans too. Now, flies have about 13,000 genes. But DSCAM is a bit complex. Genes code for RNAs, which can be mixed and matched and spliced in various ways. To make a protein, the body "reads" the RNAs, not the genes. So the RNA is the "barcode" that is produced from copying something on the gene, then this barcode is spliced and diced and glued together in numerous ways before being "read" by the biologic "barcode reader" to turn it into something (a protein) that can participate in life. Turns out that the DSCAM gene can be spliced in almost 40,000 different ways, each of which translates to a unique unique "barcode" that specifies a unique protein. That means that one gene---DSCAM---can produce 40,000 unique proteins. This is THREE TIMES the total number of genes the fly has in the first place!
Well, this situation is dramatic, but not unique. So there are estimated to be about 25,000 genes (40,000 tops), but there are at least 300,000 (maybe more like 500,000) different proteins. But the complexity hardly stops there. Because once a protein gets made, it gets structurally embellished in zillions of ways, all of which can affect its function. And, proteins can exist in partial structures, partly degraded for example. And by far most proteins don't operate by themselves, but instead in small or big groups, not infrequently mixed with many other types of proteins, all affecting one another. And all these things are flying around interacting with one another and modifying one another in numerous ways on the fly. The genome at best provides a freeze-frame surrogate of this complex, highly dynamic process that cannot possibly account for the interactions or time-dependent processes that are the very substance of life.
The point is this: the foundation of biology (and pathologies such as cancer) lies in proteins, not genes. Genes tell you next to nothing about what proteins are even around, much less what they might be doing, with who, and for how long, and how they are being controlled. Probably a majority of genes in the body take up space but never do anything at all for the entire length of the individual's life---they remain "silent."
Remember, individual genomes are about 99% similar in all humans. And, of the 1% of the genome that is variable, most of this doesn't code for anything at all, and has no known function (so-called "Dark DNA"). Of the genetic variability that is involved in coding for proteins, about half of that doesn't change the protein one bit (conservative or silent substitutions). And humans are about 98% similar to mice, and about 50% similar to bananas. True story.
So while the NCI has made some bold steps forward and is to be congratulated for really taking some risks that potentially have huge payoffs, I can't quite see the rationale behind this one, or at least I can't see enough of a rationale to justify that kind of expenditure.
Lee Hartwell is no dummy, so I hope he can channel the money in ways that really maximize the benefit.
But if I were head of the NCI, I wouldn't invest half that much in genetic/genomic anything. I've oversimplified a lot here; it's not that genetics/genomics are barren grounds, just that there's so much more fertile ground to plow out there.
T |
