Hacker News Reader

Comment by veddox | Hacker News Reader

So these guys come along, casually expand a well-established code and test it under precisely one small condition. And when this one test gives them some nice data, they say: "Hey, our stuff is better than everything all of nature has ever done!"

A very intellectually stimulating endeavour no doubt, but I expect some more tests before I would call this good science. Claiming that "the new additions appear to improve the alphabet" is simply extrapolation to the nth degree. [1]

Oh and by the way, when the article claims that

> "the three-biopolymer system may have drawbacks, since information flows only one way, from DNA to RNA to proteins"

that is not correct either. For more information, read up on epigenetics.

[1] Note that this quote comes from the article, not the original paper. The original paper is not quite as cocky (at least not in the abstract, but I don't have full access).

View on HN

I can't help but notice the irony in your comment. So this guy comes along and having read one article, says "I can't call this good science."

A very basic summary of molecular biology of the cell:

DNA - library of blueprints, basically instructions on how to build proteins

RNA - copies of blueprints you take out of the library to build proteins so you don't expose DNA to unnecessary hazards

protein - catalyzes reactions so the cell can do stuff, including make new DNA when replicating.

A fundamental conundrum exists when it comes to evolution of this mechanism. DNA is needed to build proteins, but proteins are needed to catalyze the reactions necessary to build DNA. It's a chicken and egg problem... what came first?

When it was discovered that RNA can catalyze certain reactions, presumably because of their slightly higher chemical reactivity, it suggested a way out of this conundrum. What if life originated with RNA only, where RNA acted as both the hereditary and catalyzing machinery?

The problem with this RNA World hypothesis is that the reactions that current RNA can catalyze is very limited. But what if, at the origin of life, when nature could experiment, the genetic alphabet that RNA could play with was bigger, potentially leading to expanded capability? Scientists like Benner have worked for three decades to try to answer this question.

So, your characterization of "comes along, casually expand" a well-established code and claims it's better is grossly unfair.

I am well aware of the basic molecular biology of the cell, as well as the RNA world hypothesis.

All due respect to Benner for his work - my comment was rather too pointed, I'll concede that. Nonetheless, I am always wary of too much theory being induced from too little data.

Benner's experiment shows that an expanded genetic code can form molecules that show greater chemical functionality in a given situation than that of natural DNA molecules. Now a quote from the abstract:

> This suggests that this system explored much of the sequence space available to this genetic system and that GACTZP libraries are richer reservoirs of functionality than standard libraries.

Already he is starting to extrapolate when he starts talking about the extended libraries in general. The Quantamagazine article then goes on to say:

> In other words, the new additions appear to improve the alphabet, at least under these conditions.

That is true, but for a rather narrow definition of "improve", and a very narrow set of conditions. The result is that the superficial reader goes away thinking "they've made a better DNA".

> The problem with this RNA World hypothesis is that the reactions that current RNA can catalyze is very limited. But what if, at the origin of life, when nature could experiment, the genetic alphabet that RNA could play with was bigger, potentially leading to expanded capability?

Transfer RNA currently have a few additional bases, probably to do same things that are impossible/difficult with the standard 4 bases. See: https://en.wikipedia.org/wiki/Transfer_RNA#Structure

>that is not correct either. For more information, read up on epigenetics.

Epigenetics does not violate the central dogma of molecular biology. Copy-paste from wikipedia:

>These epigenetic changes may last through cell divisions for the duration of the cell's life, and may also last for multiple generations even though they do not involve changes in the underlying DNA sequence of the organism;[5] instead, non-genetic factors cause the organism's genes to behave (or "express themselves") differently.

Note that the original quote is a common misunderstanding of the central dogma, and is incorrect (for example reverse transcription RNA --> DNA is common in nature).

The central dogma was invented by Crick, and states that once sequential nucleic acid information (either DNA or RNA) has passed into protein, it cannot be recovered. This has never been violated.

Amusingly, it was none other than Crick's pal Watson who popularized the incorrect version via his college textbook, and it is this incorrect version that is regularly announced to have been 'disproven'.

Thanks for your comment. The word sequential is absolutely key in this context.

"And when this one test gives them some nice data, they say: "Hey, our stuff is better than everything all of nature has ever done!""

https://en.wikipedia.org/wiki/Dunning–Kruger_effect

Steven Benner is an expert in the field. Dunning-Kruger does not apply at all.

It looks more like he's trying (EDIT: or his funding foundation is trying) to talk up his research, which is understandable.

It's not him, it's the simons foundation which funds his work.

I wasn't referring to anyone's expertise in the field, but to humanity's understanding of a five billion year old intelligent system. What we've done so far is impressive, but we've really only scratched the surface. It's a bit early to claim superiority.

We actually do know something about why the genetic code might have only four codons, and other aspects of its structure. It gets into combinatorics and search.

http://www.ncbi.nlm.nih.gov/pubmed/9732450

Reply on news.ycombinator.com