Saturday, December 22, 2007

The Human Genetic Variation "Breakthrough"

 
"Human Genetic Variation" is the scientific "breakthrough" of 2007, according to Science magazine. I have a problem with science journalism when science writers misuse the word "breakthrough" but that's topic for another posting [Breakthrough of the Year in Science].

In this thread I want to discuss the actual choice made by Science editors. Elizabeth Pennisi describes the choice in the lead article of this weeks issue [BREAKTHROUGH OF THE YEAR: Human Genetic Variation].
Equipped with faster, cheaper technologies for sequencing DNA and assessing variation in genomes on scales ranging from one to millions of bases, researchers are finding out how truly different we are from one another.
There is some truth to this statement. It's true that the details or the amount of genome-wide of variation are being added to the databases. But is it true that we only realized for the first time in 2007 that humans are different from one another?

Of course not. We've known about massive variation in populations since the the 1960's [The Cause of Variation in a Population]. We've been using DNA fingerprints to identify criminals for more than 15 years. Think about it. Would DNA fingerprinting work if we weren't all different from one another at the level of genome sequence?
The unveiling of the human genome almost 7 years ago cast the first faint light on our complete genetic makeup. Since then, each new genome sequenced and each new individual studied has illuminated our genomic landscape in ever more detail. In 2007, researchers came to appreciate the extent to which our genomes differ from person to person and the implications of this variation for deciphering the genetics of complex diseases and personal traits.
We're familiar with the writings of Elizabeth Pennisi so it shouldn't come as a big surprise that she makes statements like this. She seems to be remarkably deficient in her knowledge of scientific background and history.

It is simply not true that "In 2007, researchers came to appreciate the extent to which our genomes differ from person to person." Real scientists have known and appreciated that fact for decades. It's part of understanding junk DNA, Neutral Theory, and the importance of random genetic drift.

The true part of the statement is that by mapping more and more specific examples of variation we can do some experiments that we couldn't do before. This is an advance in technology but not an advance in our understanding of the extend of human genetic variation.


16 comments :

  1. While human polymorphisms at the nucleotide level are not recent news, the amount and extent of copy number variability is. In particular attention was sharply focused on CNV in late 2006 with the rise of tiled array comparative genome hybridization (CGH) for example:

    Global variation in copy number in the human genome. Nature. 2006 Nov 23;444(7118):444-54.

    A comprehensive analysis of common copy-number variations in the human genome. Am J Hum Genet. 2007 Jan;80(1):91-104. Epub 2006 Dec 5

    That entire genes and megabase-sized genomic regions differ in copy number between human individuals was not illuminated by the Human Genome Project, owing to the invisibility of this kind of high-fidelity (90+%) large-scale (larger than BAC insert) repetition to standard sequence contig assembly technologies.

    We now also know that recombination-driven CNV makes human genomic restructuring much more dynamic than was previously appreciated. This was slowly entering the literature over the last several years (championed by James Lupski at Baylor) with increasing elucidation of "genomic disorders" caused by non-allelic homologous recombination (NAHR), but the extent of segmental duplication in the human genome ( http://projects.tcag.ca/humandup/ ) potentially subject to this class of rearrangement was also only recently appreciated.

    We knew all about nucleotide-level human genomic variability, but such broad sub-chromosomal structural genomic variation was unanticipated. I think Science's choice is spot-on.

    ReplyDelete
  2. I think the genuinely interesting thing is just how few genes the human genome in fact has; about 25000.

    It IS genuinely interesting that there are great differences between individuals in the number of genes and their genomic positions. This could account for a diversity of effects beyond that of simple possession or not of single gene.

    What about this that the entire genome itself is actually quite small?

    If indeed gnomes are much smaller than computer programs, I think that is a comparison that should be quite sobering to those who use informatic metaphors as a way to understand biology.

    ReplyDelete
  3. anonymous says,

    While human polymorphisms at the nucleotide level are not recent news, the amount and extent of copy number variability is.

    Really? I remember teaching about this back in 1990. The examples I used were variations at the β-globin locus and in the MHC region.

    Being well aware of the extend of duplications in Drosophila, I wasn't the least bit surprised to find a similar situation in humans.

    Apparently, you were so shocked to discover this that you consider it a real breakthrough in our understanding of genomes, right?

    ReplyDelete
  4. Sanders says,

    I think the genuinely interesting thing is just how few genes the human genome in fact has; about 25000.

    Elizabeth Pennisi agrees with you. Unfortunately you are both revealing your lack of knowledge of earlier predictions [SCIENCE Questions: Why Do Humans Have So Few Genes?].

    ReplyDelete
  5. No, few genes fits in nicely with what I expected. I don't doubt that many people could have predicted this result. But now the genome has been actually sequenced.

    You would fool yourself, badly, if you thought this small amount in genes does not take people by surprise. Since most people believe the genotype= phenotype fallacy, they expect phenotypic complexity to relate to genomic complexity and thus would expect the human genomes to be larger than those of a nematode.
    I think it i good to rub this data in people's faces. People are mostly "gene happy" and think "a la Watson", that every difference between people is becuase of a gene.

    ReplyDelete
  6. I think the overall lesson should be: genes are less important than we thought they were.
    Sorry to all ya gene-happy people out there.

    ReplyDelete
  7. For instance, even guys as good as Mayr and Jacob have favored an informatic analogy of organisms and computers at almost literal value. The genome is the program, the blueprint.

    It may seem obvious to state that organisms do not work like computers that execute gentetics programs, but it may be necessary anyway.

    ReplyDelete
  8. I think the overall lesson should be: genes are less important than we thought they were.

    Commit non sequiturs much?

    ReplyDelete
  9. I'm out on a limb here, but having fun, hehe.

    I think the accumulated data challenges the popular notion that the primary sequence of DNA contains the "information" to construct an organism, much like running a program in a computer. Truth is, there just isn't much information storage capacity there, in the primary sequence.

    Of course, the philosophical reasons are even clearer. The coding and decoding of information are defined by the user, and only make sense to the user, in the context of the questions he/she wishes to resolve. An abbaccus provides a nice example: assigning values to different colors is what allows us to make simple operations that we know will deliver the correct result.

    Th way DNA works in the in the organism actually has nothing to do with the ways human use machines to process encoded information.

    ReplyDelete
  10. I agree with some of that (at least as I interpret what you're saying), it's just that none of it has anything to do with the fact that humans have fewer genes than some people predicted. ;)

    Irrelevant but interesting historical note- from my days working in a Drosophila lab in the 80s, I remember similarly overenthusiastic "estimates" of the number of fly genes, which were rightly snickered at by classical geneticists familiar with more realistic estimates derived from saturation mutagenesis screens on specific chromosomal segments.

    ReplyDelete
  11. The true part of the statement is that by mapping more and more specific examples of variation we can do some experiments that we couldn't do before. This is an advance in technology but not an advance in our understanding of the extend of human genetic variation.

    well, the technology has driven much insight into the extent of human genetic variation, and how it is organized--see the hapmap phase II paper from this year, and the copy-number papers. yes, people knew snps existed, and the beta-globin locus is a classic cnv, but technology now allows unprecedented analysis of the scale of variation and their evolutionary patterns. and yes, this is an advance in understanding.

    ReplyDelete
  12. It's not only the absolute number, but the comparison across taxa. It takes about the same amount of genes to build a nematode than to build a human (I guess is this another of those "humbling" scientific moments for humanity).
    And, our TOTAL genome, non-coding included, does not fill up a blank CD! It doesn't matter if you believe in the wondrous regulatory capacity of non-coding dna. Again, pretty humbling.
    And then of course we have the complete lack of correspondence between genome size and phenotypic complexity.

    Genes (DNA) are overrated. To me , in the light of this data, there is no way that we can say that genes or DNA in any way contain the encoded "information" to build an organism.

    ReplyDelete
  13. Steve Labonne says,

    Irrelevant but interesting historical note- from my days working in a Drosophila lab in the 80s, I remember similarly overenthusiastic "estimates" of the number of fly genes, which were rightly snickered at by classical geneticists familiar with more realistic estimates derived from saturation mutagenesis screens on specific chromosomal segments.

    Interesting thing about Drosophila is that most of the early estimates were too low! The estimates from the 1970's were less than 10,000 genes while the actual number is about 12,000!

    ReplyDelete
  14. Yup, there were estimates all over the map, including drastically low ones as you say (I seem to remember the number 5,000 being bandied about quite a bit back then.)

    Of course if there were arthropod scientists proud of the unique specialness of their own phylum they probably would have been shocked when the number turned out to be "only" around 12,000!

    ReplyDelete
  15. Mats,
    What a surprise...an ID proponent that also believes in noah's arch? I think Mats is about to deliver us a sermon. Reach out for your bibles, people. Its the only book a "real" scientist needs, no? Hehe.

    The bible story of noah and the flood is not consistent with the current state of knowledge of geology and biogeography, even though it was once the established, received view. Yet already in the XVIII century, ID believers like cuvier were proponing at least several deluges or cataclismic events. Your noachian explanation contradicts everything we know about geology and the natural history of planet earth. Maybe you have an intelligent geology" theory in mind?

    Indeed, having all the fossil taxa living at the same time and saying that they are the evidence of the global flood is about the level of goofiness I expect from a completely typical dogmatic bible-bonker young earth creationist.
    IIf Mat won't even admit he is a creationist, how can we expect him to be portraying himself sincerely as an "old earth" creationist? I see absolutely no good reason WHY Mat would NOT be a young earth creationist. After, all, all h has to do is deny scientific fact when they do not fit his scheme.
    We' re curious mat. What convinced you that the earth is actually older than around 6000 years?

    ReplyDelete
  16. Sanders - I agree with you that DNA sequence is probably overrated. But your argument that the number of genes is "small" and therefore insufficient to encode complex organisms is totally useless. Why is 25,000 small? What makes you think 25,000 genes is too few to make very complex objects? Liquids consisting of only a single substance can give rise to very complex structures - think for example, of snowflakes.

    Or, can you tell us how many genes exactly are necessary to make a functional human body?

    I don't think you or anyone can answer that question accurately at this point.

    ReplyDelete