The research paper is not about the "evolution" of the amino acids. The research paper claims that a selection of 10 early amino acids (instead of the current 20) is better to produce random proteins. It makes sense that the initial living cell used a safe list of amino acids so most of the time it makes a protein that doesn't suck. Now with better building enzymes we can use an extended list and get better proteins. The important part of the research paper is that some of the easiest to produce amino acids are not good candidates for the safe list.
From the article:
> Now, a new study in the Journal of the American Chemical Society suggests that evolution began long before life emerged and that proteins swirling around in the primordial soup selected for preferable traits. In other words, there was a sort of Darwinian selection effect taking place (even prior to the emergence of life) among the un-living proteins and amino acids in the primordial soup.
I don't agree the research paper says that.
> * Stephen Fried, a Johns Hopkins chemist who co-led the research with scientists at Charles University in the Czech Republic, said in a statement. "We're describing the events that shaped why that ancestor got the amino acids that it did."*
That is what I understand. The research paper is about amino acid selection in the early living things.
> So by testing which libraries had the best solubility and structure, the researchers could conclude that proteins were evolving and driving natural selection before even forming living things. Over time, the proteins that were the best shape for biochemical processes were incorporated into the fundamental cycles of life.
I don't agree the research paper says the first sentence.
The second sentence is very very unlikely, because there is no known way at the cell level to look at a useful protein and reverse engineer it and get the DNA or RNA code to produce it.
---
I'm a fan of the RNA word hypothesis. https://en.wikipedia.org/wiki/RNA_world It's still not proven, so take the rest of my comment like an opinion.
The idea is that RNA can store info and be copied so it can replace DNA, and it can be a catalyst so it replace proteins. So you can start with the tools to build one type of molecule instead of three.
One of the problem is the transition from RNA to DNA, but they are quite similar and we still see transference of info in both direction, so it's not unimaginable.
The transition form RNA to proteins is harder. The translation in the cell is not straightforward and there are many proteins and pieces of RNA to make it possible. Another problem is that if the initial cell has good RNA catalyst https://en.wikipedia.org/wiki/Ribozyme , how to get an equivalent protein catalyst https://en.wikipedia.org/wiki/Enzyme
One of the current unproven ideas is that the transition can be done in steps:
1) 100% RNA catalyst
2) RNA catalyst with a short chain of amino acids to help
3) A catalyst with a part that is RNA and another part that is a long chain of amino acids or a short protein
4) An enzyme that need s a coenzyme that is a short RNA chain.
5) 100% enzyme.
And there are many intermediate steps. The transition can be done step by step removing one RNA base and adding one amino acid (not necessary to replace one by one simultaneously) I think we have seen 1), 4) and 5). Ribosomes can be interpreted as a huge 3). I'm not sure we have seen a 2), but it's would be not surprising.
---
Back to the article, if at the beginning of the transition from RNA to proteins the initial cells used a safe list of only 10 amino acids in the brute force search, it may have been slightly easier.