The Institute for Creation Research has a charming little magazine called “Acts & Facts” that prints examples of their “research” — which usually means misreading some scientific paper and distorting it to make a fallacious case for a literal interpretation of the bible. Here’s a classic example: Chimps and People Show ‘Architectural’ Genetic Design, by Brian Thomas, M.S. (Note: this is not the peer-reviewed research paper implied by the logo to the left — that comes later.) The paper is a weird gloss on recent work on CNVs, or copy number variants. Mr Thomas makes a standard creationist inference that I have to hold up for public ridicule.
At one point in their Genome Research report, the investigators referred to the chromosomal arrangement of CNVs as “sequence motifs or architectures.” “Architecture” logically suggests an architect. Although these researchers have encountered architectural (designed) features in DNA, they hold fast to their belief that nature itself is the architect, even though scientists have not discovered a plausible mechanism or observable example.
It would make sense that a Master Architect fashioned these attributes–some unique to chimpanzees, some unique to humans, and others quite similar in both–in genomes so that each group could better survive the different diseases it might encounter. These genetic “architectures” provide evidence for purposeful design and are more consistent with being the products of an all-knowing God as the Bible describes, rather than the products of indiscriminate nature.
Whoa. Scientists use the word “architecture” all the time in describing genes and genomes, and it doesn’t mean anything like what Thomas thinks it means. When you look at a gene, for instance, you can describe it in terms of its structure: the arrangement of introns and exons, for instance, or the location of regulatory regions. Because it is not a random gemisch of noise, but actually exhibits arrangements of regions relevant to its function, it’s fairly common to refer to it as having an architecture. It does not imply an architect; it implies an observed pattern, nothing more.
An architecture does not imply intent or purpose, but they often imply a history. The pattern described — that chimps and humans share some common structural elements in their genomes — is better described as evidence of common ancestry than of well-designed function. An intron, for instance, is a piece of random, usually useless DNA inserted into the middle of the sequence of a gene that must be excised from RNA before it can be used to make a functional protein. It’s a little piece of garbage that must be cleaned up before the gene product can do its job. That a human and chimpanzee gene has identical introns is an example of an architecture, true enough, but it is of a shared error. Some all-knowing god—he seems to be consistently making the same mistake.
This next part is particularly interesting, because Darwin’s insight is staring the author right in the face, and he doesn’t quite get it.
Some specifications of CNV design are that they are regions of DNA that are variably repeated, are often found in zones that are copied and spliced, and contain immune system (and few other) genes. They do not contain core genes, which are essential to basic cell function. If such genes were housed in an area that was more vulnerable to volatility and change, the survival of the cell itself would be at risk.
How does the cell “know” how to protect itself in this manner? According to the evolutionary model, it is the result of non-directed natural selection. According to the creation model, the Creator wisely placed core genes in regions of chromosomes that are much more stable. Conversely, it seems reasonable that the same Creator placed other genes in more volatile regions of chromosomes for the purpose of allowing more variety and survivability in succeeding generations.
The original research he is discussing does point out that there are some areas of the genome that are hotspots for copy number variation, and others that are not. These CNV regions tend not to include certain sensitive genes, genes that are not tolerant of changes in dosage. If those genes are varied, the organism tends to die, as Thomas himself notes. Wouldn’t the surprising thing be to observe lethal variants in living organisms?
But no, he instead leaps right into a common creationist error: assuming that the only way this unsurprising result could happen is if cells “know” something, that they plan to avoid the circumstance. They don’t! They die, so they don’t show up in the populations of living organisms we study!
It’s a shame Thomas is so unaware, too, because CNVs are actually looking very interesting. So forget the garbled creationist gobbledygook for now — let’s explain what a CNV is.
Here’s a typical human karyotype. These are the chromosomes from an ordinary diploid cell, like most of the cells in your body, which is why they come in pairs. You have 23 different kinds of chromosomes, each with a unique complement of genes, numbered from 1 to 22, with one other special case, the X and Y chromosomes that differ by sex. Let’s just focus on chromosome 1 to keep it simple.
So here’s chromosome 1. Most of your cells have two copies of this chromosome, which means that they have two copies of each gene on that chromosome. For instance, near one end of the short arm of chromosome 1, there is a gene for an amylase — an enzyme that helps you digest sugars. I’ve guessed about where that gene would be located, and have drawn a red bar there to indicate the amylase gene.
Again, note that this cell has two copies of the amylase gene. If it had only one copy, it would probably produce less amylase; if it had more copies, the cell would be able to pump out more amylase. This is the dosage of the gene. For some genes, dosage is very important. Having too many or too few copies could affect the rate of metabolic or developmental processes, for instance; too few, and maybe the lower levels would choke off growth, too many, and inappropriate processes might run wild at the expense of others.
We know that humans can’t tolerate overdoses of some genes. The best known example is Down Syndrome, caused by having an extra copy of chromosome 21. Chromosome 21 contains several hundred genes, and instead of being in duplicate, each are in triplicate in someone with Down Syndrome. This difference sends development down subtly awry pathways, producing individuals with a suite of problems.
We also cope poorly with reduced numbers of copies for some genes. Cri-du-chat syndrome, for example, is caused by a deletion of part of chromosome 5, so these individuals have only one copy instead of two for all of the genes present in that area, although the key gene may be just one, CTNND2. Not having enough copies of the gene product CTNND2 leads to errors in nerve growth, producing microcephaly and mental retardation.
But let’s return to chromosome 1. What happens if we have an extra copy of the amylase enzyme gene? As it turns out, nothing that we can detect. Maybe these individuals would have saliva that was especially good at breaking down french fries, but no physiological evidence of any difference has been found yet. They’ve just got a spare, or maybe cellular processes regulate enzyme expression so no real difference occurs, or maybe they do make a little extra enzyme, but digestion isn’t significantly effected. Unlike variations in the dosage of a gene like CTNND2 that controls gene growth, variation in the dosage of a digestive enzyme don’t seem to be as critical to the human organism.
This is what CNVs are all about: scientists have examined the genome of different individuals, and found that there is a significant amount of variation in copy number for some regions, and the part of chromosome 1 that codes for an amylase is one of them. What that means is that we can find people who have chromosomes like this:
He has a duplicated copy of amylase on one of his chromosomes — two identical versions of the same gene, with another copy on the other chromosome, for a total of three amylase genes. And it’s not just amylase: there’s a whole block of DNA in that neighborhood that is duplicated to varying extents in different people, so that this locus varies between 150 kb and 425 kb long in different individuals. There are also individuals who lack one copy of the gene, carrying a deletion on one chromosome, like so:
None of these differences cause any known, visible phenotype, which is the interesting thing about CNVs. We all have variations in the dosage of some genes scattered throughout our genomes; maybe you have four copies of some gene, your spouse has 2, your best friend has 3. Studies have found on the order of a dozen different regions with variant copy number between any two people — which is almost certainly a gross underestimate.
CNVs are not easy to detect. You can’t just look at the superficial phenotype and spot them, nor can you ask simple yes/no questions about the forms of the genes present, since the extra copies are often identical in sequence. Instead, the typical strategy is to extract small slices of the genome, find copies of the gene of interest, and ask if all the copies have the same neighboring DNA sequences. It’s not trivial and it’s easy to miss duplicates, but it’s easy enough to find lots of examples of variation between individuals in human populations.
That’s what’s particularly interesting about CNVs: they are another potential source of variation. We’re used to the idea of different forms, or alleles, of genes in a population, and how selection can work to remove one form or another. CNVs are subtle variants in gene dosage rather than gene sequence, and who knows? They may be responsible for some useful phenotypes that we have difficulty measuring.
The paper specifically addressed (or misaddressed, or just plain missed) by the creationist journal suggests that there is selection for some differences in dosage. The researchers compared the genomes of 30 humans and 30 chimpanzees, looking for CNVs within and between groups. They found some general observations, such as that CNVs tended to cluster around regions that were prone to large scale duplications. That suggests that some caution in interpreting shared CNVs is necessary — you don’t know whether an arrangement of genes in one area is shared between a chimp and a human is because they inherited those copy numbers from a common ancestor, or because they shared a hotspot for duplication with a common ancestor.
You can look for consistent copy number differences between species and regions of copy number variation within a species, and ask whether there are differences that have been fixed within each lineage, and where and in what kinds of genes those differences lie. There are some patterns. The copy number differences that show evidence of fixation in evolution (which may be a result of selection) are associated with genes involved in inflammation and with cell proliferation. These may be correlates of changes that led to our differences — the genes regulating cell proliferation are especially provocative, since we do have one organ, the brain, that is dramatically different in size and cell number.
However, the interpretation of the evidence must be completely different from what the creationists claim. What we find are patterns of random variation in gene copy number within human populations, which is interesting in itself, but also shows that the ‘architect’ had to have been remarkably nonchalant about the specification of his creations; and also that that variation can be part of the fuel for the evolution of differences between species. Only a creationist could read a paper describing the existence of a reservoir of chance variation and turn it around to claim it means that there was a god who placed the gene numbers with specific intent for specific purposes.
Freeman JL, Perry GH, Feuk L et al. (2008) Copy number variation: New insights in genome diversity. Genome Res 16: 949-961.
Perry GH, Yang F, Marques-Bonet T et al. (2008) Copy number variation and evolution in humans and chimpanzees. Genome Res 18: 1698-1710.