All Humans Are Mutants, Say Scientists 309
Hugh Pickens writes "In 1935, JBS Haldane, one of the founders of modern genetics, studied a group of men with the blood disease hemophilia and speculated that there would be about 150 new mutations in each human being. Now BBC reports that scientists have used next generation sequencing technology to produce a far more direct and reliable estimate of the number of mutations by looking at thousands of genes belonging to two Chinese men who are distantly related, having shared a common ancestor who was born in 1805. To establish the rate of mutation, the team examined an area of the Y chromosome which is unique because, apart from rare mutations, the Y chromosome is passed unchanged from father to son so mutations accumulate slowly over the generations. Despite many generations of separation, researchers found only 12 differences among all the DNA letters examined. The two Y chromosomes were still identical at 10,149,073 of the 10,149,085 letters examined."
What about non-humans? (Score:3, Interesting)
I get 450 mutations per generation (Score:5, Interesting)
3600 mutations total
8 generations in 200 years
450 per generation
5 in protein coding section of genome
Weird Headline (Score:5, Interesting)
Rather than making me think that all humans are mutants, this made me think: Wow, over a runtime of 204 years, the DNA copying process has an accuracy of 99.99988%, or an error rate of only 0.00012%.
I think we'll be hard-pressed to replicate that level of awesomeness in computers anytime soon.
Re:Quality reporting (Score:3, Interesting)
SMBC [smbc-comics.com] is completely accurate on this count.
Yep, it's obvious that we're all mutants, how else does evolution happen? The bbc seems to have missed the point, which to me is that they've now got a decent (they claim) estimate of the rate of mutation. This is, however infinitely less interesting than the bbc title.
A more interesting variation should be done (Score:4, Interesting)
Re:Weird Headline (Score:5, Interesting)
Uh, we do all the time.
The diploid human genome is 8 gigabases. Each base encodes 2 bits of data. That is 4GB of data per genome. Let's say that a gamete is produced after 1000 generations of cells from the fertilized egg (no idea what the right number is, but I suspect that the true figure is lower). That means that 4TB of data is being copied, with an error rate of 450 bits.
If I want I can set up two 4TB raids on my server at home (assuming I had more disk space), and issue the command dd if=/dev/mdx of=/dev/mdy bs=1M count=4000000. Then I could do a diff on the two volumes. I'd be shocked if they had any errors at all.
These kinds of error rates are actually not all that uncommon with computers.
Now, the 204 year bit sounds impressive, but it isn't like a piece of DNA lasted 204 years without any decay. Instead it was copied repeatedly over that time. If I copied that 4TB hard drive once every 25 years (generation time) onto a brand new drive (assuming that you could keep making them compatible) I don't think that getting the data across 200 years without any bit-flips is really that tall of an order. Sure, technology will change, but that really is a different matter, and I doubt that any commodity computer technology used in the next 200 years will do any worse than what we have today.
Re:Weird Headline (Score:2, Interesting)
Wow, over a runtime of 204 years, the DNA copying process has an accuracy of 99.99988%, or an error rate of only 0.00012%.
While I agree that the level of change is reasonably slow, I think you've taken the conclusion a bit too far in inferring the observed rate of change matches transcription accuracy.
The reason I would be cautious about extending observed mutation rate to infer transcription accuracy is that there is likely to be significant selection bias, similar to how "old furniture" always appears to be great quality (because anything that isn't great quality is in a landfill). Any fatal mutations would never progress and therefore can't be detected by this method. Thus, the 0.00012% is a (very) loose lower bound on the transcription error rate.
To follow your computer analogy, it's like saying a program running for 204 years only produces a wrong answer 0.00012% of the time *that it produces an answer*. What you may be missing is the 50% (making up a number) of the time that it dumped stack because a bounds check failed due to an error.
Re:Article title seems stupid to me (Score:3, Interesting)
Sounds like they simply confirmed with real data what before was simply believed to be extremely likely.
Re:Article title seems stupid to me (Score:4, Interesting)
Yeah, the "all humans are mutants" angle doesn't have much to it. Of course we're mutants insofar as we're the product of evolution, and evolution requires mutation. Without mutation, you wouldn't get new genetic differences to be weeded out or passed on. So yes, life is a mutation and we're all mutants.
It will be interesting now that we could be able to sequence your DNA and your parents' DNA, figure out exactly what mutations you have (if any) from the previous generation, and possibly know what those mutations do. Maybe in the future we'll be able to map all of our genetic family trees in detail, figure out when traits were introduced, and see what patterns emerge. Maybe those random mutations aren't so random.
Fun article, though mutation is sporatic not const (Score:1, Interesting)
Fun article. Not statistic but fun nonetheless.
For those who don't know systematics, the Y is used precisely because it changes little over time. Ditto for mitochondrial DNA (which you get from Mom alone). The idea is to determine relatedness via differences and commonality. The y changes slowly therefore more commonality, so easy to make a tree. The 'non-coding' regions are being found to contain transcription factors, controlling factors, and lost bits of mutation or viruses (you have retroviruses in your DNA fyi).
Re:Um... statistically significant? (Score:3, Interesting)
But it's restricted to two people, or not even that, it could be just one different ancestor. Maybe one's grandfather was exposed to radiation, or mutagenic chemicals.
Y chromosome is special (Score:4, Interesting)
The Y chromosome [wikipedia.org] doesn't get to recombine [wikipedia.org], so measuring the mutation rate of the Y chromosome only gives us a limited understanding of mutations in general.
Lack of recombination means you don't get to measure mutations that consist of genes being brought together for the first time in an individual. It also eliminates entire classes of accidental mutations. On the other hand, it removes the opportunity for some types of in-cell DNA repair [wikipedia.org].
Furthermore, the Y chromosome is less interesting than most. It contains very few working genes, precisely because it is not subject to the most important [wikipedia.org] DNA repair mechanism of all: sexual reproduction.
Re:Why? (Score:3, Interesting)
Even if it was ammo, would you really listen to someone who believed that humans were formed from dust or a clot of blood and continue to believe the parlor tricks of old mystical texts?
I wouldn't, but some of my friends, relatives do (In addition to several of our lawmakers). I also do not avoid being an evangelist for what I consider rational thought. Therefore, I do care what BS is flowing through the collective minds of the religious crowd. It is akin to me knowing a lot more about homoeopathy than several of my acquaintances who actually believe in its efficacy. These people actually feed their babies sugar pills (I do not see how placebo effect can help babies even if that is the one part of homoeopathy that works) instead of treating them as they should.
The ignorant are not the problem though. It is the irrational minds who corrupt the ignorant minds that we need to be wary of.
Re:Weird Headline (Score:5, Interesting)
Now, the 204 year bit sounds impressive, but it isn't like a piece of DNA lasted 204 years without any decay. Instead it was copied repeatedly over that time. If I copied that 4TB hard drive once every 25 years (generation time) onto a brand new drive (assuming that you could keep making them compatible) I don't think that getting the data across 200 years without any bit-flips is really that tall of an order. Sure, technology will change, but that really is a different matter, and I doubt that any commodity computer technology used in the next 200 years will do any worse than what we have today.
Actually, it's more than copying the drive once every 25 years, it's making a copy of data on the drive many times each day -- some where around the 100,000th copy of the drive randomly choose a copy to keep and start the process over again. With that kind of usage on a drive, the failure rate (let alone error rate) will be _much_ higher.
Re:Ammo for the ID nutjobs? (Score:3, Interesting)
As this experiment shows, you might have accumulated a few hundred single nucleotide polymorphisms- differences at one base pair- in the lineage from your great-grandfathers to you. However, so much shuffling of the genetic deck occurs in each generation's gametes that, as may be obvious to you, two people (siblings, for example) can be closely related but display very distinct traits. The reason why you'd want to focus on the Y chromosome if you wanted to isolate the mutation rate is that it doesn't undergo all of this shuffling; you probably only have a maximum of one (there are a few XYY males)and it passes down patrilineally with only random mutations to change it. Those two men tested could well have very similar Y chromosomes, but otherwise be genetically very different.
I would argue that there is a survivorship bias in studying mutation in the Y chromosome, though. There aren't many genes on the Y chromosome, but the ones it does have tend to be critical for producing healthy, fertile males. It might be the case that mutation rates that might be tolerable on other (somatic) chromosomes produce completely inviable offspring when they occur at that rate on the Y.
Re:Article title seems stupid to me (Score:4, Interesting)
Maybe those random mutations aren't so random.
This is complicated and not really worth going into in depth here, but a major technique of mapping species divergence and establishing when they diverged is through mapping the number of mutations that have shown up in non-expressed DNA. The mutation rate of DNA is fairly well known (it's largely a function of the precision of the enzymes that duplicate DNA, the DNA polymerases and their error-correction fidelity, which varies between different DNA polymerases.) There are some wrinkles in that many mutations don't survive -- they're lethal -- and that's why some parts of DNA are referred to as 'conserved', because those sections can't tolerate changes. There are genes involved in vision, for instance, that have something like a 0.3% difference between insects and humans. But sections that aren't critical, or aren't used at all, chunks of old viruses that got spliced in and don't do anything, accumulate errors. Taking a quantitative diff of two DNA strants gives you a number that is proportional to how long ago the species diverged.
Re:Article title seems stupid to me (Score:3, Interesting)
It will be interesting now that we could be able to sequence your DNA and your parents' DNA, figure out exactly what mutations you have (if any) from the previous generation, and possibly know what those mutations do.
It would be very unlikely for you to have no (germ cell) mutations from the previous generation. It's fairly easy to arrive at an order of magnitude estimate of the number of mutations that are uniquely yours. I'll save you the math, but that number is about 10. Only about one in 25,000 people has no mutations of their own.
Of those 10 mutations, many are in non-coding areas of DNA and tend not to cause a problem. Some will inactivate a gene, which is why we have multiple copies of every important gene, except those on the X and Y chromosome in men. (Ever wonder why men are more susceptible to genetically linked diseases?) Some could change the structure of a protein, which may or may not be a problem depending upon whether the new protein has similar activity to the precursor, whether it has no activity, or does something harmful. Cases where a mutation will be advantageous are pretty rare.
Of course, if one of your 10 mutations were instantly fatal you never would have been born. That may be the reason that the majority of pregnancies don't even implant successfully, and of those that do, many don't last for more than a few days or weeks. Some mutations may kill you later in life, even after you reproduce. However mutations to coding regions of DNA tend to be moderately fatal and will probably get weeded out in a few generations. Your surviving great great grand children are unlikely to have any mutations to coding DNA that came from you.
Re:Article title seems stupid to me (Score:3, Interesting)
Yea and so is the summary... The very next line says that of those 12 mutations, 8 of them occurred in the lab. Only 4 occurred naturally (which btw confirms JBS Haldane's conjecture).
What I'd like to know is WHEN those 4 occurred. Roughly 200 years since these fellows last shared an ancestor, say 10 generations. Yea.. it happens gradually but it DOES happen so... somewhere in the sequence granddad-dad-son at least one mutation occurred for someone. What was that mutation like? Did the kid not look like the father? Was it completely unnoticeable (by eye)? Was it a random cosmic ray to the balls that caused it?
Re:Weird Headline (Score:1, Interesting)
The cells that produce gametes (the germ line) are set aside early in development, and their proliferation is carefully controlled from a set of stem cells. So the number is greater than 25, but much less than 10^5 copies per day.