Hacker News Reader

Comment by WClayFerguson | Hacker News Reader

People also underestimate how easy it is for life to get started by accident. Let's say you need a short segment of DNA to randomly pop into existence, capable of nothing but self-replication. What are the odds?

Here's my best analogy: If you take one mole of "random" Rubik's Cubes (the number of molecules in a glass of water), then search for 'Solved' ones you'll find 512 are magically solved, merely by the power of large numbers. The ways molecules can 'snap together' are similarly 'finite', just like a puzzle.

Now think how many glasses of water are on earth. A replicating molecule would have existed probably on the FIRST DAY earth had cooled enough to form water.

If we knew what the shortest sequence if DNA is that can replicate, we'd be able to mathematically state the probability of it existing in any 'random' length of DNA of that length as 1/(4^N) where N is DNA length (number of AT/GC pairs).

View on HN

> If we knew what the shortest sequence if DNA is that can replicate

Take it as 200 bases and you won't find another life in the observable universe. Avogadro's number is no match for exponential growth.

First approximation - 50 base pairs of RNA? (Unless I am reading the supplement wrong)

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3943892/

"The present study describes the directed evolution of replicating RNA enzymes that operate with an exponential growth rate of 0.14 min−1, corresponding to a doubling time of 5 min. Each parental enzyme can give rise to thousands of copies per hour, and each of these copies in turn can do the same, all the while transmitting molecular information across the generations."

Thanks for that link! That's amazing. I had no idea that work had ever been done.

According to my math 10 million moles of random sequence 50-base pair RNA segments would be required to ensure a correct 'Hit' on a replicator, assuming only one of those 50-long sequences can replicate. 4^50=10^30. 10^30/Avogadro=10^7. Assuming each enzyme is the size of a water molecule, that would be about one large swimming pool of water.

This makes the assumption that all the RNA molecules of 50 bp length are "clean" chains of nucleotides, without any chemical modifications that break the replication process. This is highly unlikely if the RNA is not produced by existing enzymes, but by more or less random chemical processes.

It's purely the math probabilities I was getting at above. If you take the number of molecules in a large swimming pool, that is the same number of 50-long RNA molecules it would take to statistically cause at least one RNA to accidentally have any specific pattern of length 50.

In actual evolution this number of RNA molecules can be thought of as 'diluted' across the entire oceans, and not magically sitting in single pond.

https://nebiocalculator.neb.com/#!/ssrnaamt 10 million moles of RNA of length 50 turns out to be a large amount of RNA... but then again there was like a billion years for life to develop, all around the world.

All around every world in the universe. If life had emerged on planet 5 around star 1778465 in galaxy M33 instead of earth, then people on that planet would be having the same thoughts.

> Let's say you need a short segment of DNA to randomly pop into existence

But do you? If elements A and B tend to form compound A-B, and compound A-B can attach a further set of A and B, forming B-A-B-A which under certain circumstances will split into A-B and A-B, you have basic reproduction, and perhaps the occasional presence of element C will expedite the process. Game of life. In my [very] layman's view of the world, given a billion years or two, some such infinitely simple process should be enough to get the ball rolling, as long as it can somehow produce an ever so slightly errorprone copy.

That's right. Once you have a replicating strand that's modular (linearly able to glue end-to-end) it can also recombine randomly with other completely different replicators to see if the "offspring" can replicate. We know for a fact major sections of DNA are shared among lots of different organisms, so we know this did happen a lot during evolution.

Going from a short segment of DNA randomly popping into existence, capable of nothing but self-replication to living organisms, carrying a tremendous amount of intelligence, those are two very different things.

To make an analogy, what you describe would be only the electronic chips or the logic gates of a computer. There are a still the CPU, RAM, and the whole software stack from the OS that runs our bodies (and mind you, there is a different one for each species) to the yet unknown abstractions from which human general intelligence emerges.

I think people underestimate how difficult it is for all the above to get started by accident.

Once the process is started with simple life evolution just causes complexity as a rule. I think people who can't believe evolution made us from "nothing" have never really tried to understand the time frame in which it happend, which is a really long time.

Humans can't conceive of large timescales nor large numbers. That's why I use my Rubik's Cube example. A single glass of water contains enough molecules that if they were each a random Rubik's cube then there's a mathematical certainty that 512 of them will be perfectly solved (per glass).

For molecules that are a linear chain of only 2 possible items in the chain, it becomes a 'brute force search' of a 'puzzle space' to find a chain that 'does something like computer code', and the earth has the computing power to solve that brute force problem in 30 seconds, not 30 million years.

But this is more of a proof how insanely big a mole is, rather than the fundamental question. Also, Rubik’s cube possibilities is not that high. Try to create each permutation of a standard deck of cards (hint: there are less atoms in the universe than the number of permutations)

About the 'card deck' combinations: That's why we can be virtually certain there are no life forms in the universe that consist of a chain molecule (as it's version of "DNA") where there's more than a handful of different types of molecules along the chain.

Nature is statistically guaranteed to find the simplest patterns first and that's why our life is built on 'binary' (two kinds of 'cards' AT/GC). Life on all planets will likely use binary "information storage", especially because of it's modularity (i.e. it can be cut-n-pasted together in different patterns)

Life is not built on ‘binary’, there are 4 bits. All four nucleotides take part in the encoding (reading happens on one thread only).

And since life was presumably started with RNA, we are still talking about 4 base elems

It's perfectly legitimate to say life stores information using only two "types" of things (AT or GC) which can be oriented one way or another, which is a "binary choice of the type". I explain it that way because it reaches a wider audience (easier to visualize)

However there's always at least one stickler who wants to quibble that the orientation is also a "binary bit" of information. Congratulations, you're him today!!

I did think about that, that’s why I added the second paragraph, namely that what about RNAs? (Since presumably they precede the DNA world) They can self-bind (I don’t know the proper nomenclature in English), but by all means they have 4 bases.

There's two kinds of base pairs, is identical to there's 4 kinds of bases.

RNAs are a single chain. Your point makes no sense.

A code with a 4 letter alphabet would be 2 letter, if it happened to use a parity “bit” or other error correction info?

I never said RNA was built on two types of molecules. I said DNA was.

Wouldn’t that imply that we’d find independent genesis of replicating molecules frequently even now?

If we found a sequence of DNA on some sample, and another identical sequence in that same sample, we still wouldn't know replication is happening.

However there may be slightly more requirements like the sequence has to be found in an oil film (bubble) up until it can form a skin/membrane, to protect itself, so even if it was happening on earth trillions of times every day no one would be there to discover it, and other life on the planet with a head start (bacteria, etc) would consume it before it made more 'progress' by evolution.

In the early planet there was nothing to "eat up" life as it was starting. Everything was on an equal footing.

There's strong evidence that all life on this planet can trace back to a single strand of DNA, which evolved exactly once in our billions of years.

For example, if I understand this correctly, a random jumble of DNA as you describe would have a 50/50 chance of twisting to the right or twisting to the left. And yet all life on earth uses DNA that is "right handed" twisting to the right.

The fact that life is right-hand twist DNA probably means left-hand was unsuccessful due to chemistry either in itself or in the environment, because an inversion of a molecule (chirality, twist, etc) means it will behave completely differently chemically. So left-hand twist can exist, but not necessarily in a way that can build life...even if it's able to initially replicate.

And then calculate the odds that the given DNA (though the RNA hypothesis seems more plausible) is in sufficient conditions like inside a lipid-bilayer with sufficient nucleotides for a reproduction - which is more than likely thermodynamically not a stable state to be in, so add another factor with an e^(-x) expression.

AT and GC are such simple molecules we can assume wherever we find 50 of them in one place (first replicator) you're going to have not just 50 but an abundance of them in that place too. And since that's what the replicator needs to replicate it's literally sitting in a bath of it's own "food" with no predators around, so it will grow like crazy and immediately. lipid-bilayer might be needed to find a pattern for "cells" to form, sure, agreed. And in all this 'base' counting math the statistics is the same regardless of whether it's DNA or RNA.

Thermodynamics would beg to differ about your first sentence. Also, I’m not trying to disprove your point, since life did emerge in some way and I’m not fond of unscientific explanations. But it is naiv to assume that life is a definite outcome given an Earth-like situation with a primordial soup.

> Thermodynamics would beg to differ

It's already experimentally proven you can mix up N, H, and O and it spontaneously forms RNA. We've even found entire RNA and sugars on meteorites. It's not rare in nature, it's abundant and self-emergent from the raw atoms themselves.

I replied in terms of your first sentence. Molecules doesn’t concentrate spontaneously without active energy expenditure from someone (putting salt in a pot will have salt uniformly distributed everywhere).

Now, either there were some naturally formed impermeable areas with an abundance of base material (yet again, I’m not saying life emerged because the UFOs created it), or we are talking about many orders of magnitude less statistical chances. I am willing to believe that many, not individually extraordinary thing had to happen at the same time, that made life possible - but that pretty much makes their combined presence extraordinary, which is my claim.

(Some form of underwater volcanic caves or whatever could sound like a possibility, especially because last I checked, starting sticked to a surface with certain catalytic activities is a plausible theory. Especially due to the reduced statistical chances due to a 2D surface instead of a 3D one! What I have a bit of trouble seeing, is how did organic materials form - I am aware of the famous experiment with lightning bolts that created ammonia and some other chemicals, but it would only slowly increase their uniform distribution in the oceans. Perhaps some chemical reaction with gases that made them accumulate? )

AT and GC never concentrate spontaneously. Brownian motion makes them 'find' each other and immediately snap together like tinker toys when they find a partner that fits. That's why you need warm temperatures, and it's why all chemical reaction rates are proportional to temperature. Even in a solution with 0.00001PPM reagents, reactions take place rapidly when warm.

> People also underestimate how easy it is for life to get started by accident.

That's well known. It is the key driver for the global contraceptives market, which generated revenue of $28,175 million in 2015 and is estimated to reach $43,812 million by 2022.

Reply on news.ycombinator.com