Qt45: A small polymerase ribozyme that can synthesize itself

Qt45: A small polymerase ribozyme that can synthesize itself(science.org)

117 points by ppnpm 79 days ago | 21 comments

cyberax 79 days ago |

It's interesting, that we're getting down to the sizes of self-replicating RNA that realistically can form by a complete accident.

Getting this sequence by random chance out of a pile of nucleotides is a 1 in 2^90 chance. That's around 1.2*10^27 or just around 20000 moles! Not at all an impossible number.

jjk166 79 days ago | |

Note that the Bennu asteroid sample had approximately 5 nanomoles of nucleotides per gram, meaning 20,000 moles of nucleotides could be delivered by a single 4 million ton asteroid, which if it were a spherical carbonaceous chondrite would be about 183 meters in diameter. An asteroid about that size impacts earth roughly every 36,000 years, and that mass of meteor material falls to earth each century.

If primordial earth's oceans had nucleotide concentrations comparable to Bennu, then there would be about 10^39 nucleotides in the ocean.

dekhn 79 days ago | | |

I don't see any reason for the source molecules to come from space. We already know that nucleotides will spontaneously form and polymerize in conditions consistent with the early earth, and a meteorite origin just moves the source of those nucleotides elsewhere but doesn't answer how they formed.

nkrisc 79 days ago | | |

If raw materials isn’t the bottle neck for life every where, then what might it further down the line between oceans full of nucleotides and life? The oceans themselves?

littlestymaar 79 days ago | | |

I wonder how much nucleotides present in asteroids in space survive the massive heating in the atmosphere and the crash though.

adgjlsfhk1 79 days ago | |

it's much more than 1 in 2^90. this specific 45 base sequence was found by random search which strongly implies that the odds of finding such a protein are much much higher (i.e. >2^-30) since the experiment probably only searched a couple million proteins

cyberax 79 days ago | | |

(these are not proteins, these are RNA sequences)

And it's likely that there are even smaller replicators but probably not much smaller. And there are likely more configurations among these 45 base-combinations that can self-replicate. It's also more likely that the first replicators were arrangements of multiple smaller molecules that can amplify themselves, and not just one large molecule.

But even this longer sequence is well within the realm of synthesis by pure chance. RNA molecules can grow base-by-base, so a random walk model should eventually produce it.

Kotlopou 78 days ago | | |

Actually, the search here is chemical -- you make a bunch of random small RNA sequences, then chemically replicate those that do what you want, and after a few cycles of this you find just a few sequences dominating the population. The pools can easily reach 10^14 examples and more, because you don't need a separate container/experiment for each one, but do it all together in one batch. The keyword is SELEX, a pretty cool idea: https://en.wikipedia.org/wiki/Systematic_evolution_of_ligand...

Paedor 79 days ago | |

20000 moles of length 45 nucleotides is 306 metric tons? Spread out over millions of years, that does seem completely feasible.

java-man 79 days ago |

2009 paper: Self-sustained replication of an RNA enzyme [0].

[0] https://pubmed.ncbi.nlm.nih.gov/19131595/

rcyeh 79 days ago | |

Amazing! The 2009 Lincoln & Joyce paper you cited catalyzes one bond per hour on average. (Doubling time = 1 hour, but only one bond between oligonucleotides needed to double.)

OP's Gianni et al 2026 paper connects 45 nucleotides, taking 72 days (1700 hours) to yield 0.2%.

The latter effort is like drawing the whole owl.

That is incredible patience. Without access to the full article, I read only the abstract. I wonder if they used simulations to narrow the candidates?

java-man 79 days ago | | |

https://pmc.ncbi.nlm.nih.gov/articles/PMC7618777/pdf/EMS2125...

Noumenon72 79 days ago | | |

I don't know if ChatGPT translated your rate for the 2026 paper right, but it says you mean:

* 99.8% of the starting molecules fail to form the desired product

* The ones that do, form 0.026 bonds per hour

pcrh 78 days ago |

The pay wall doesn't permit much access to the details of this. A better summary is available at the UK's MRC website, link below.

https://mrclmb.ac.uk/news-events/articles/bridging-the-gap-f...

panarchy 79 days ago |

Quinozyme

guax 78 days ago |

We haven't finished migrating from Qt5 and there is a v45 already? Jeez.

dekhn 79 days ago |