My aunt was good at puzzles. They shoulda called her.
May 21, 2019 4:25 PM   Subscribe

The human genome has never actually been complete, because reading our highly-repetitive centromeres has been "like putting together a puzzle of the Sahara Desert." Paper.
posted by clawsoon (14 comments total) 12 users marked this as a favorite
 
God damn I want me some nanopore reads so bad.

I work on insects with modestly sized genomes and I stick to the transcriptomes anyway because the assembly is so much easier if you can predict protein coding regions easily. I got my start on salamanders, though, and salamander genomes... They make the human genome look like a little kid's cake walk. Only lungfish have bigger genomes. And why? No one fucking knows why. Why do they have all these repetitive elements, tandem repeats for literally miles, no one knows, possibly it's just because they're really really bad at moving long distances so their effective population sizes are small so they just don't get the unnecessary genetic material selected out? But I don't really buy that. I kinda think it's just contingent, they just happened to get a million DNA viruses that left themselves implanted in the code.

Hey whose Y chromosome was it? Are we still just sequencing Craig Venter's DNA for the alignment's sake? I see that they're sequencing from BACs but where'd those clones come from I wonder?
posted by Made of Star Stuff at 6:50 PM on May 21, 2019 [7 favorites]


Not to brag, but, um, I’ve got a GridIon, and we just passed acceptance testing. So, you know, time to order more discs for the SAN!
posted by wintermind at 6:54 PM on May 21, 2019 [3 favorites]


Only lungfish have bigger genomes. And why? No one fucking knows why

Hmm. I had heard that it was lilies that had the longest DNA and that the going theory was that it was a mechanism to isolate viral DNA from replicating. Makes sense for salamanders and lungfish too b/c swampy dirty bottomfeeder environment prob full of viruses. But I could see how proving that would be an issue.
posted by sexyrobot at 7:14 PM on May 21, 2019


Hey whose Y chromosome was it?

A man in Buffalo who probably responded to this ad 22 years ago. Not Craig Venter.
posted by grouse at 7:50 PM on May 21, 2019 [2 favorites]


Thanks to you I'm still stuck on why pufferfish don't have retrotransposons, clawsoon.

And now this!
posted by jamjam at 9:31 PM on May 21, 2019


This is a cool achievement, thanks for posting it!

I've played with nanopore sequencing a bit (just a single MinIon, so I'm envious of wintermind's new toy) and got some promising data for sequences that are short but tricky for Sanger sequencing. But disappointingly the bosses aren't keen for us to sink the resources into properly optimising it for our projects, so it's rattling around in the back of a drawer in the lab.

Thanks to you I'm still stuck on why pufferfish don't have retrotransposons, clawsoon.

Wait, what? This review says that they do. And, brilliantly, TIL that there's a clade of retrotransposons in fish called "sushi".
posted by metaBugs at 2:25 AM on May 22, 2019 [1 favorite]


Yep, pufferfish do have retrotransposons, they're just really well controlled. Apologies if I gave anyone a wrong impression! The difference is that while we have millions of copies of various retrotransposons, pufferfish only have a few copies. Our genome is ~50% retrotransposons (mostly broken), while pufferfish have... is it 1%-ish? 5%-ish?
posted by clawsoon at 4:07 AM on May 22, 2019 [1 favorite]


About salamanders and lilies: As I understand it, doubling of chromosomes is much more common in plants than in animals, most likely because the development of animals is more of a Rube Goldberg mechanism in which one thing going wrong will cause a whole lot of other things to go wrong, while plant development is looser and more robust. (Not related to chromosome doubling, but a good example: If a tree gets an extra limb, or ten extra limbs, it's no big deal. If a mammal gets ten extra limbs, it's got a problem.)

...or so I thought, but it looks like some animals go polyploid pretty much whenever they feel like it:
Examples in animals are more common in non-vertebrates such as flatworms, leeches, and brine shrimp. Within vertebrates, examples of stable polyploidy include the salmonids and many cyprinids (i.e. carp). Some fish have as many as 400 chromosomes. Polyploidy also occurs commonly in amphibians; for example the biomedically-important genus Xenopus contains many different species with as many as 12 sets of chromosomes (dodecaploid). Polyploid lizards are also quite common, but are sterile and must reproduce by parthenogenesis. Polyploid mole salamanders (mostly triploids) are all female and reproduce by kleptogenesis, "stealing" spermatophores from diploid males of related species to trigger egg development but not incorporating the males' DNA into the offspring. While mammalian liver cells are polyploid, rare instances of polyploid mammals are known, but most often result in prenatal death.
Huh.

Back to sequencing, this raises one question I'm curious about: Trying to sequence a salamander with 12 sets of chromosomes seems ridiculously difficult, but the same basic problem is there even with animals like us that have 2 sets of chromosomes. How do you deal with the fact that we've got two sets of polymorphisms, not just one?
posted by clawsoon at 5:32 AM on May 22, 2019 [1 favorite]


Made of Star Stuff: And why? No one fucking knows why. Why do they have all these repetitive elements, tandem repeats for literally miles, no one knows, possibly it's just because they're really really bad at moving long distances so their effective population sizes are small so they just don't get the unnecessary genetic material selected out? But I don't really buy that. I kinda think it's just contingent, they just happened to get a million DNA viruses that left themselves implanted in the code.

I have a half-baked theory about this which is probably wrong: If you have some active, not-completely-controlled retrotransposons, it's useful to have a lot of mileage of inactive, useless retrotransposons, because that gives the active ones lots of harmless junk* to land on without doing any real damage. If a retrotransposon were to slip through the pufferfish's defenses, there's a good chance it would land on something important and cause serious damage, because nearly everything in the pufferfish genome is important. In a human or a mouse, on the other hand, it's more likely to crash into a retrotransposon junkyard that makes no difference.

If you drop a bomb on a random spot in Siberia, you're probably not going to kill anybody. If you drop a bomb on a random spot in Singapore, different story.

If this probably-wrong idea is correct, it seems that both states would be somewhat self-sustaining: Any pufferfish which lets a retrotransposon land will likely have a major genetic problem and be selected out of the population, while any human who starts cleaning up junk retrotransposons is reducing the safe landing area for active retrotransposons. (How often to retrotransposons successfully duplicate in humans? This 2009 study says between 1-in-20 births and 1-in-200 births. 40 million years ago, it was one every birth.)

*Note: I say "junk" very, very loosely. I know that there's a debate about how much of the "junk" is actually junk. The "junk" might be biologically or structurally important in a way that has driven our evolution and/or is important in gene expression. People in this discussion who are much more qualified than me can weigh in on that.
posted by clawsoon at 5:56 AM on May 22, 2019


This is going to make me late for a meeting, but now I want to calculate how many miles of DNA we have. This study says we have 3.72e13 cells, minus (from Appendix B) 2.63e13 red blood cells with no DNA, gives 1.09e13 cells with DNA. At ~2 meters of DNA per cell, that's about 24 billion kilometers of DNA. That's only one-sixth of the way to the sun. Pffft. Hardly anything.
posted by clawsoon at 6:23 AM on May 22, 2019 [1 favorite]


...times 7 billion people is... 17 million light years??? Did I do my math correctly? Would all the DNA in all humans really reach to the Andromeda galaxy 7 times???
posted by clawsoon at 6:39 AM on May 22, 2019 [1 favorite]


that's about 24 billion kilometers of DNA. That's only one-sixth of the way to the sun.

? Huh? No. The sun is only 150 million km away. Pluto is (on average) about 6 billion km from the sun...so (6Bx2=12B) our total DNA is ~2x the orbit of pluto.
posted by sexyrobot at 7:15 AM on May 22, 2019 [1 favorite]


On my bicycle ride here I was thinking the same thing. Someone should double-check my math from the start, though, since I did it in a hurry.
posted by clawsoon at 7:36 AM on May 22, 2019 [1 favorite]


This reminds me of the truism about how if you lined up someone's blood vessels along the edges of a football field, you'd probably get arrested.
posted by metaBugs at 7:40 AM on May 22, 2019 [8 favorites]


« Older Like Christmas every day   |   Jazz - Out of the Blue Newer »


This thread has been archived and is closed to new comments