Doug Axe Tweets a Paper on Non-Coding RNA

1 Like

What do you think @art?

What is he trying to say exactly? Just in my brief internet search knowing very little about the topic it seems non-coding RNA is an active area of research within the community of scientists who accept evolution. It doesn’t appear to be news then that non-coding RNA… well does something:


Not going to be able to digest the whole paper at the moment, but it probably warrants mentioning that this is a completely theoretical paper by a physicist on how computation could be done using noncoding RNA. And the quote about noncoding RNA not being “junk” comes from the discussion and is a summary of other literature, not a conclusion of this research.


The first two sentences of the introduction make me groan.

It seems, upon a very quick reading, that the author is proposing massive amounts of intermolecular RNA-RNA associations. I have to go back and read other papers, but I don’t think this model is consistent with in vivo RNA structure probing studies.

Figure 9 makes no sense to me - it looks like the author is suggesting that RdRP does not end up making dsRNA. Figure 10 reminds me of the PolIV - PolV system in plants. I haven’t really thought to superimpose RdRPs into that sort of mechanism. That is something to think about.

I don’t think I will wade through the details.


Quote mine maybe?

It is a dog whistle comment, I believe, riffing on the theme that “mainstream science thought non-coding DNA is useless junk, whilst ID predicted it would be functional; look how ID was right.”


Of course there are many varieties of noncoding RNA, including the long-known tRNAs and rRNAs in addition to some of the more recent sorts. It’s clear that some of them are functional, but it’s also nearly as clear that most of the rarely transcribed sequences are junk. Just being transcribed doesn’t demonstrate function. Typical ID rhetoric: look, here’s a pseudogene that has a function, therefore all pseudogenes are functional; look, here’s a transposon that has a function, therefore all transposons are functional. Nothing to see here, folks.


Can any IDer out there please explain the reasoning behind the claim “ID predicts non-coding DNA will have function”. How could ID possibly predict that without insider knowledge of the Designer’s intentions and mechanisms? Like all ID “predictions” this one actually is a postdiction. Science discovers some interesting new factoid and the IDers all sing in unison “ID predicted that!”.

1 Like

This is probably right and Behe agrees with you.

I wouldn’t call it that, exactly. I think it accurately reflects the sentiment of the author (Deutsch). And there is some justification for the statement in the papers Deutsch cites. So I think it is fair game to quote his summary of those papers. In a more formal academic setting, one would need to cite those papers directly, but some informality is expected on Twitter.

At the same time, it is a convenient quote for the purposes that @CLAVDIVS describes.

1 Like

Upon further review …

The first two sentences I noted before:

The overwhelming majority of transcripts in the human genome produce non-coding RNA (ncRNA) and these have been under intensive investigation in recent years [1–6] which has revealed many functions. However, research to date has still only scratched the surface of the mechanisms involved with these transcripts.

I guess that, when someone tells the RNA community that they haven’t done much, feelings may run hot. However, when I recollect any number of seminars about diseases, almost all of which begin with brash claims about how the particular disease of interest, whatever it is, is the most devastating to affect humanity, then I guess I can pass this opening off as acceptable hyperbole.

What is harder to accept is that the author, Dr. Deutsch, parlays some mathematical models into some ideas about RNA dependent RNA polymerases in human cells, and how these might help in performing some sorts of computation. The problem is that human cells do not possess RdRPs; the citation Deutsch (ref. 20 in the paper that readers here can view at the link above) uses to claim otherwise actually doesn’t show this. (There is an interesting side story behind reference 20, but it’s too long and off topic for this thread.). So the entire exercise makes no sense, since the molecular connection doesn’t exist.

But the paper may inspire some novel ideas. While human cells do not usually have RdRPs, cells infected with RNA viruses do, since RdRPs are viral replicases. I have no idea what Boltzmann machines and the Hopfield model are, but if they can lead one to more accurate variants of Figures 9 and 10 that pertain to viral replicases, maybe something interesting can come out of efforts like this. Who knows. (I know that I don’t, but there are some computational biologists in this forum who might fancy some idle speculation about this sort of thing.)

Regardless, what annoys me is that, while human cells do not possess RdRPs, plants do. And they are known to play central roles in the small RNA regulation. It would have been nice if Deutsch could have gotten his “system” correct.

Why would Axe point to this paper? Again, who knows. If he thinks there is some ID-friendly aspect to this, then I believe he is mistaken.


So what is the ID position on junk DNA anyway? That there is none? A little? How much?

They never come out and make an explicit prediction that give an acceptable range of how much junk DNA there should be in the genomes of any organisms. They always just vaguely hint that if some entity isn’t junk, this is somehow vindication of a never explicitly stated ID position on junk DNA.

1 Like

Behe does not have a position. Others claim that ID predicts function. I agree with Behe that there is not predictive model within the ID frame work.


Yeah I would have to agree with that too. But that just makes it a bit mysterious why many ID proponents still make it a point to argue about it.


There are some problems with respect to internal consistency when it comes to ID and junk DNA. On one hand, ID supporters argue that function in sequence is highly specified and changing the sequence will most likely damage that function, not help it. On the other hand, about 90% of the human genome accumulates mutations at a rate consistent with neutral drift. In the end, one of these concepts has to be wrong.


Sternberg made a clear and explicit prediction here (speaking to Darrell Falk):

All I am asking for is a prediction, such as “90% of these DNA letters is superfluous” (“or 79.5% of the RNAs are nonsensical”). Since he also said “almost certainly” in the above statement, he must have a figure in mind. So I say pick a number, any number…But to be a good sport, I’ll show my prediction: All of the expressed 88.5% of our DNA has diverse roles in our development.

The “expressed 88.5%” he mentions comes from the ENCODE project, if I recall correctly.

1 Like