Hacker News Reader

Comment by rflrob | Hacker News Reader

Paragraph 4: “Slowly, slowly, slowly, the terminology of ‘junk DNA’ [has] started to die,”

Rest of the article: continues to use the term "Junk DNA".

I think more than any other group, programmers can understand the gut reaction of calling code that they don't understand "junk".

Normally it means "poor quality" though. I think "junk DNA" is more analogous to "dead code". Not all code labeled "dead" is actually unreachable/unused.

Well we do have "comments" that are actually interpreted by tooling.

Everyone doing biology professionally already knows the non-coding DNA has function. It’s only logical since we know that resource-intensive mechanisms don’t survive the evolutionary pressure if there’s no benefit for them to exist. So it’d be wasteful to replicate 98% of DNA if it’s not functional. Keep in mind it’s not only replicated, there’s error-correction as well.

it's wasteful to replicate non-functional DNA, yet there are organisms whose genomes are 10X-100X larger than any others (literally ten copies of a genome), and they do just fine (https://en.wikipedia.org/wiki/Paris_japonica). The cost of DNA replication isn't that strong a functional selection afaict.

Sounds like "non-functional DNA" might in fact have a function, hm?

Probably not in the context of the example I gave; that's more likely to just be a totally random genome duplication which will eventually be functionally selected away.

Non-functional DNA is to all DNA as source code that doesn't call I/O is to all source code.

You're conflating the size of a genome with the amount of DNA that is non-functional.

I'm not conflating anything. Those details are scientifically conflated (IE, they are functionally related) in a very interesting way. For exmaple, new genes probably arise from gene or genome duplications.

DNA that doesn't encode proteins is not necessarily non-functional.

Yes, I know that. It's unclear what aspects of those duplications are "functional but not coding for genes, enhancers, promotors, or other well-known and understood coding and regulator elements", and which are "functional because if you remove it the organism dies but we can't explain what the cause is".

Everything in my terminology is entirely correct within the current mainstream understanding and it encapsulates much of the ongoing confusion about what "role" non-coding functional elements play, and how they play it.

See http://cryptogenomicon.org/encode-says-what.html and https://www.cell.com/current-biology/comments/S0960-9822(13)... for the strongest arguments against ENCODE's more wide definition of functional elements, as well as https://www.pnas.org/content/110/14/5294 and https://www.mun.ca/biology/scarr/MGA2_02-10.html

I am of the opinion that nearly everybody in this field is overly dogmatic in their views of how genomes work at the macro scale, and that these squabbles demonstrate that clearly.

For more context, please understand that as a researcher in this field, much of what I say may not make sense to casual readers. I decided it's better to think about minimal organisms, than ones that have 500 copies of the ribosome (http://citeseerx.ist.psu.edu/viewdoc/download?rep=rep1&type=...)

And yet we probably don't know why they have so many copies. Maybe it's a crude method of regulating the ratio of activation on genes. If a few genes are either on or off, and you want to have ratios other than 1:0 1:1 or 0:1, maybe independently activating N of the M copies can achieve that. Just speculating as to how multiple copies might be useful.

You can't prove that they are "extra copies" until you understand pretty much everything about them and their interactions with everything else.

While I agree in principle, it does look like in most cases, these 100X large genomes are just due to random events that haven't been selected away. Often, what you see looking at evolutionary processes at a point in time, you want to find a "reason for why it's this way", when the reason is "it was random chance". Compare deinococcus radiodurans, where there's a whole mechanism for exploiting multiple copies of the genome. Plants wiht 10X genomes don't seem to exploit that, afaict.

If it didn't give an advantage it shouldn't have become dominant (that all individuals of that species have that) or not?

that's now how evolution works. evolution tolerates enormous amounts of things that don't 'give advantage', and 'advantage' is a fairly subjective term anyway. Any number of possible events could have occurred to lead to the scenario we see. Population bottlenecks often lead to strange outcomes.

It’s always random chance. What’s non-random is why it’s kept in the genome. I’m no plant expert, but genome duplications could be linked to increased fitness due to amount of resulting protein product, right?

If the organism is pressured, the non-coding DNA slowly fades. There are some parasitic organisms (eg Nasuia deltocephalinicola) that have highly compacted DNA: no introns, no regulatory regions, no transposable elements and even overlapping genes.

Introduction: Introduces term.

Early text: Details evidence that term is misguided.

Late text: Continues to use early term so as to provide a consistent interface to readers. Might use 'scare quotes' if term is particularly egregious but still conserves familiarity of terminology.

What would you have them do instead? Naming stuff is hard. The authors get their point across quite nicely without also trying to coin more terminology.

Reply on news.ycombinator.com