LianaLi

Anybody read the article about the Proteome in the current issue of Scientific American? I did, I should have bought it, but I started mumbling at the stupidity of the companies before I even finished reading the thing.

Quibble 1:
Three years.

That's the target date some companies have set, for completing the human proteome. That includes knowing all the proteins, their functions, and structure.

Anyone besides me think that's a LITTLE too optomistic a completion date? The Genome project took what, a decade almost? And they want to finish the protiens, which strike me as much more complex than DNA, in three years? Yeah right. Not to mention I think they're shooting themselves in the foot with my next quibble.

Quibble 2:
Copyrights and patents.

Companies that have figured out certain sequences and functions for protiens aren't sharing. Or, they will, for a fee. When they do complete whatever protien their working on, they run immediately to obtain a patent for it. Excuse me? Where's the paper? The articles? The publications? No, no, no, they want to protect their stuff with a patent so they can sell it later. And yet, they want to complete the whole human sequence of protiens in three years. That's just absolutely brilliant.

IMHO, the one of the accomplishment of the human genome project (HUGO) was its ability to share its research. All their findings are publicly accessible. Its major competitor, Celera, was able to complete the human genome using techniques TAKEN from HUPO. Celera couldn't have completed the genome without the shotgun technique, taken directly from the methods they were using at HUGO. Because of this sort of open collaboration, they completed at the same time.

And the various companies working on the proteome want to patent all the protiens they find, so no one else can use it. It seems to me, they are shooting themselves in the foot this way, because they won't be able to collaborate. Or, other researches can't use the same methods for other protiens, because it's under a patent. To me, that just seems like a huge setback, a detriment to collaborative work. Maybe I got the wrong impression from the article about just exactly what they're trying to patent.

Otherwise, three years is a HIGHLY optomistic number, since not too much is known about protiens. Yes, HUGO did alot sequencing the genome. Look! We know the sequence for 3 billion base pairs. That's nice. What do they do? We aren't even too sure what the functions for all the junk DNA are, and it consitutes so much of the genome. It's useful for pinpointing disorders, genetic predispositions to cancer, and/or stipulating what could be the cause for types of cancer, but it goes hand in hand with knowing the protiens the DNA codes for. And locking that information under a patent just doesn't strike me as the wisest thing to do. Lock treatments away under a patent for money, but not information.


-Liana

because I can't SPELL

[ April 10, 2002: Message edited by: LianaLi ]</p>

lpetrich

I agree that finding the function of every single protein in our bodies is a grandiose goal. And a much more difficult one than simply sequencing a genome. Consider that the Escherichia coli genome is very well-mapped, yet nearly 1/5 (899 out of 4732) of its Open Reading Frames (ORF's) code for unknown proteins. Source: <a href="http://genprotec.mbl.edu/misc_inf.html#Summary" target="_blank">http://genprotec.mbl.edu/misc_inf.html#Summary</a>

Also, this insistence on getting patents I agree is misplaced. However, there is the interesting question of how all this work is to be funded. Companies that finance some biomedical work may not be happy with letting competitors leech off of their work. And after the dotcom bust, venture capitalists may not be very eager to fund companies whose business plans state, in effect, "we're not sure how we are going to make money off of this thing", which was common during the dotcom bubble.

But there had been a big biotechnology bubble in the late 1980's and early 1990's, and there might be one again.

[ April 10, 2002: Message edited by: lpetrich ]</p>

The Lone Ranger

For what it's worth, keep in mind that when work was first begun on the Human Genome Project, people were predicting that it would take at least 30 years to complete, perhaps 50 or more.

No one anticipated the vast increase in computing power that became available however, and the project took only about a decade.

Still, the proposed project is gargantuan -- one that seems like it would take much longer than 3 years.

Cheers,

Michael

LianaLi

I was talking with my boyfriend about this, and he reminded me that things are not driven by pure ideals, as I would like to think, but that the almighty dollar is at the bottom of it all, driving it, funding it, allowing it to happen. So, while patenting what they accomplish can run counter to expanding research among several companies, it protects their ability to make money.

I still think it's stupid, but I bow before the almighty dollar.

-Liana

lpetrich

I've thought some more about this question, and I've come up with additional thoughts. The business-politics issue of patents seems difficult to resolve, so I'll consider the technical aspects.

Once the genome sequence is well-established, the next thing is to search it for Open Reading Frames (ORF's). These are segments of the genome that code for peptides, amino-acid changes. One complication is the presence of introns; I'm not sure how well they can be recognized.

Detecting ORF's and their introns is not the end of the story, because many proteins may have more than one peptide chain, and may have various prosthetic groups added on.

The ultimate goal is determining the proteins' function; that can be much more difficult than sequencing a genome. As an indicator of the difficulty, in the Escherichia coli genome, something like 1/3 of the identified ORF's are identified because of close sequence similarity to an ORF that encodes a protein with known function.

And that's just the protein-coding parts of the genome. The noncoding parts contain parts that are involved with gene regulation, but they also contain much "junk DNA", which lacks any clearly-recognizable function.

Many protein functions will be difficult to determine without doing experiments that would be either impractical (we are big and take a long time to grow) or unethical (anyone willing to give someone's children birth defects?) when done on human subjects.

So it will be necessary to use species that are smaller, quicker-breeding, and generally considered OK to experiment on, like small rodents, frogs, fish, fruit flies, and yeast. Evolutionary biology will give a good idea of what features are likely to be coded by related genes, though one ought to expect some surprises, like anterior-posterior, dorsal-ventral, central-nervous-system, heart, eye, and other important patterning genes being widely shared, with vertebrates being upside-down relative to arthropods.

That was Geoffroy St. Hilaire's old hypothesis, which would be re-proposed and rejected about every 20 years, for what seem like good reasons. For example, the arthropod central nervous system is shaped like a ladder, with the rung-side intersections having ganglia, while the vertebrate one starts out as a tube (can a ladder and a tube really be homologous?). And the mouth must somehow change sides.

This is something like continental drift, which had been a borderline theory for decades before paleomagnetism and seafloor spreading were discovered. Now it's well-established and even directly observed.

So there are likely (1) a lot of widely-shared gene functions and (2) lots of surprises.

[ April 12, 2002: Message edited by: lpetrich ]</p>