Oh foolish supporters of genome sequencing

Deaddog Ellington at Frodo.MGH.Harvard.EDU
Tue Feb 12 20:41:21 EST 1991


I make several general comments, which I believe pertain to most
of the material in Dr. Fristensky's reply:

(1) None of the suggested benefits requires the complete sequence
of the genome.  All of them could be realized by targeted sequencing.

     (a) For example, domains that are attached to the nuclear matrix can
easily be isolated and sequenced.  It is unnecessary to clone
them all to determine whether they are involved in higher order
chromosome structure, or in gene regulation.  In fact, a great
deal of experimentation has already been done in this area
(see, for example, JMB, 200/101, 1988; TIGS, 3/16, 1987;
for those who care, the Bodnar reference is J. Theo Biol., 132/479).

     (b) The interrelationship between speciation and repetitive
sequences is well established; see also, for example, Chromosoma,
78/137 (1980).  In fact, plants aside, it has been well-studied 
in the higher apes, including man (EMBO J., 6/1691 (1987), 
PNAS, 83/3875 (1986)).  The very fact that these studies can
be done suggests that sequencing the human genome is less than
necessary.  If you are truly serious about learning about the role of
repetitive sequences in evolution, you would do well to do comparative
sequence analysis between related organisms (the approach that has
already been taken).

2. To the extent that these are worthy goals, they could be 
carried out in other, more worthy eukaryotes (again, 
Drosophila and C. elegans leap to mind).

3. No offense, but if you were to write a grant proposal
for these ideas ("I would like to sequence the human genome to
learn about how interspersed repetitive sequences affect
speciation.") you would receive a rotten priority score.
As I have said before, I would like to see the human genome
initiative in the same pool of grants as the science it
purports to do.  I do not believe it would stand up to 
the harsh scrutiny that everyone else is having to endure.
 
4. You write:

> There are literally thousands
> of genes for which only 1-10 copies of the transcript are present in the
> cell [Okamuro and Goldberg, 1989]. It is only by obtaining a clear view 
> of what the cell does with these rare transcripts that we will have a 
> complete understanding of how gene expression results in a differentiated 
> organism.

Can you tell me exactly how the sequence of the human genome will address
this problem?  I mean, what do you do:  find which genes don't have high
abundance RNAs and then look for a function for these genes?  If so, how
do you intend to look for function?  And how is it different than just 
looking for function in the first place without the sequence of the human
genome?  One example will suffice.

Most genes are identified by phenotypes, which for the human genome will
mean RFLP mapping.  Some genes will be identified by homology to genes
found in other higher organisms that we can ethically mutate all to hell, 
like rats.  In neither case is the sequence of the human genome likely 
to make a difference in determining *function*.

I LIKE THE DARWIN ANALOGY!  Good show!

Let's carry it a little further.

Two points:

(1) Was seeing the Finches of the Galapagos the only way 
Darwin could have realized the wonders of natural selection?
(Analogy:  even if sequencing the human genome will reveal
something unknown, is this the only mechanism by which 
we can know this unknown thing?)  Obviously not:  Darwin
points out how pigeon and dog breeding are beautiful examples
of natural selection at work.  Once the insight came, the evidence
was all around him.  Wallace stumbled across natural selection
without ever going to the Galapagos.  The DATA was there; the insight
was not.  Think harder, sequence less.

(2) If Darwin had gone to the Royal Grantsmaster and said:
"I just know I can figure out how organisms change through 
time; give me a fleet of ships and I will sail
to all the corners of the Earth and collect every piece of
biological data that I can find in hopes of finding an answer."
I submit that the Royal Grantsmaster would have bunged
young Charles out on his young ass.  
     The cost of the Beagle was justified because terrestrial
exploration had proven profitable to the English time and again 
(e.g., the New World).  Why not include a snot-nosed naturalist?  

What justification is the genome initiative going to piggyback on?  
I gave you the examples of the SV40 and
chloroplast genomes:  these are what you are going to have
to take to the Royal Grantsmaster to justify your big 
project.  What have they yielded?  And why is sequencing
the human genome a better bet than a hundred other bits
of proven science (for closure, why sequence the human genome
when we can fund Roy Britten and Eric Davidson to tell us all that 
we desire about reptitive sequences?) 

Finally, we have:

> It seems likely to me
> that sequencing the genome will be cost effective, as compared to 
> hundreds of separate projects to clone individual genes.

Oh come now!  

The separate projects are seldom directed at just "cloning
the individual genes."  They are directed at "cloning and understanding
the individual genes," which is why they are much more useful than
the GeeWhiz approach of the genome initiative.  

In addition, let's say that
the average gene of interest is, oh, 10,000 bases in length.  For 100
odd genes that would be 1x10E6 bases, as compared to the human
genome size of 3.3x10E9.  Are you trying to tell me that
it will cost more than 100x more to do directed cloning/sequencing
of these target genes than it will to do the whole genome (since you
must do a sizeable fraction of the genome to be sure of finding 
most of the target genes)?  And that the concommitant knowledge of what the
genes actually does not go a long way to actually justifying that cost
(knowledge which may or may not be the product of the genome initiative)?    

> Query: Is Deaddog (Non-woof) really playing devil's advocate here,
> challenging supporters of the genome project to justify it in a better 
> way than has been done up to now? 

No, I'm really this much of a curmudgeon.  Too much is at stake.
I don't actually worry so much about starting the human genome project.
Hell, Soviet science survived Lysenko.  No, I'm worried about what
happens when we finish sequencing the genome, and Congress and the public
ask us to justify the cost.  At least with the "War on Cancer" there
wasn't a defined endpoint staring us in the face.  When that sad day
comes when the headlines blare "100% complete at a cost of N billion!"
there is going to be hell to pay.

Non-woof 

(You know, you Matrix-teers aren't really trying.  No one's even 
mentioned "introns" yet, or "heterochromatin," or "gene
dosage," or "sub-telomeric sequences," or ....  But of course 
those are all already being studied in established labs fighting
desperately for funding, aren't they?)  



More information about the Biomatrx mailing list