More Recent Comments

Thursday, August 06, 2020

More misconceptions about junk DNA - what are we doing wrong?

I'm actively following the views of most science writers on junk DNA to see if they are keeping up on the latest results. The latest book is DNA Demystified by Alan McHughen, a molecular geneticist at the University California, Riverside. It's published by Oxford University Press, the same publisher that published John Parrington's book the deeper genome. Parrington's book was full of misleading and incorrect statements about the human genome so I was anxious to see if Oxford had upped it's game.1, 2

You would think that any book with a title like DNA Demystified would contain the latest interpretations of DNA and genomes, especially with a subtitle like "Unraveling the double Helix." Unfortunately, the book falls far short of its objectives. I don't have time to discuss all of its shortcomings so let's just skip right to the few paragraphs that discuss junk DNA (p.46). I want to emphasize that this is not the main focus of the book. I'm selecting it because it's what I'm interested in and because I want to get a feel for how correct and accurate scientific information is, or is not, being accepted by practicing scientists. Are we falling for fake news?
When it was first discovered, the nongenic DNA was sometimes called—somewhat derisively by people who didn't know better—"junk DNA" because it had no obvious utility, and they foolishly assumed that if it wasn't carrying coding information it must be useless trash.

In evolutionary terms, a DNA sequence with no function is simply dead weight that gets carried along, at some cost to the organism, to be jettisoned at the first opportunity. If the sequences were not adaptively important, evolution would have kicked them out as expendable excess baggage. The fact that nonrecipe DNA continues to be part of the human and other eukaryotic genomes over millions of years indicates that there is some adaptive value to carrying the "junk baggage" along, even if that value remains unclear to us today.

In addition to various putative regulatory and structural functions, recent evidence indicates that mutations in the integernic noncoding DNA leads to a increase in susceptibility to various diseases. If confirmed, it would show a clear adaptive value to "junk" DNA.

Today, we appreciate that this is not useless junk and now call it noncoding DNA. About 80% of the DNA is known to have some activity, even if the exact activity hasn't been determined. We now usually call it the more benign dark DNA.
OMG! It looks like the proponents of junk DNA have failed to make an impression on this author.

There's a lot of misinformation in those few paragraphs, as most Sandwalk readers know. Here are a few highlights ...
  • Nongenic DNA was never called junk DNA by any knowledgeable scientist back in the 1970s or at any time since. That's misleading information that has been debunked so many times that it astounds me how any modern molecular geneticist could possibly believe it [Stop Using the Term "Noncoding DNA:" It Doesn't Mean What You Think It Means]. Some of the "foolish" scientists who taught us about junk DNA include Susumu Ohno, Sydney Brenner, Motoo Kimura, Francis Crick, Thomas Jukes, and Ford Doolittle. Apparently they didn't know any better. The list of modern "fools" is even longer and it contains the names of dozens of highly respected scientists who are experts on genomes and molecular evolution. I do not understand how any modern scientist could dismiss the work of experts in the field of genomics by imagining that they were/are stupid enough to dismiss all noncoding DNA as junk. How did that silly myth ever take hold?
  • The author has a 1950's adaptationist view of evolution. He is unaware of the fact that modern evolutionary theory can easily accommodate genomes that are 90% junk. He is also unaware of the data since there's good experimental evidence to support the idea that less than 10% of the sequences in our genome are under selective constraint
  • Five Things You Should Know if You Want to Participate in the Junk DNA Debate
  • We have known for decades that regulatory sequences are abundant and essential. They were NEVER dismissed as junk DNA. The fact that mutations in regulatory regions might cause diseases in humans is perfectly consistent with everything we known about functional DNA and it has nothing to do with junk. (I should also note that mutations in real junk DNA can cause diseases but that doesn't mean the DNA isn't junk.)
  • Once again, there was never a time when knowledgeable scientists were confused about the difference between junk DNA and noncoding DNA. A substantial percentage of noncoding DNA is junk but the most of the functions residing in noncoding DNA have been known for fifty years [What's In Your Genome? - The Pie Chart]. The idea that 80% of our genome has some sort of functional biological activity has been thoroughly debunked and discredited. Knowledgeable scientists do NOT refer to most of our genome as "dark" DNA because they are fully aware of all the positive evidence showing that most of it (~90%) is junk. It's disappointing that the majority of scientists are unaware of this evidence [Required reading for the junk DNA debate]
Let's be clear about one thing. The propagation of this misleading information about junk DNA is not entirely Alan McHughen's fault. He is merely repeating what he thinks is the consensus view of the experts—it is not his field. I might criticize him for not doing his homework before publishing this in his book but I assume that he had no idea that what he was writing was controversial.

The fault lies with those of us who are proponents of junk DNA and with science culture. Somehow, the idea that 90% of our genome is junk has failed to make an impression on our fellow scientists. Somehow, the idea that evolution includes Neutral Theory, random genetic drift, and a thorough understanding of the principles of population genetics has failed to reach the average biologist. What are we doing wrong? How can we fix it?

1. Full disclosure, Oxford declined to publish my book after I sent them a proposal.

2. Read my posts on Parrington's book at: John Parrington discusses genome sequence conservation.


João said...

It's really sad Oxford declined to publish your book. I don't why, but "there is junk in our genome" is not popular. Well, actually we probably do know why - adptationism. And naive adaptationism is not fading away. Never will.

"How can we fix it?", you ask. The only way I see is better teatching of the principles of evolution and population genetics, as you said. Things will be fixed when most biologists understand that allels can be FIXED by other means than Natural Selection. Oh, Genetic Drift, it has been too hard to be you.

Please keep working on your book. I think it will help.

John Harshman said...

Add to your complaint the idea that if a deleterious mutation can occur in a sequence, then that sequence must therefore be functional.

Lamarck said...


in principle, I sympathize with your point of view. But while I walk the sandwalk with you, allow me first to tell a little joke that Manfred Eigen used to tell in his introductory lectures:

"Once upon a time there was a shepherd who tended his sheep in a lonely place. One day a car driver stopped right next to him and asked him: "If I guess how many sheep you have, will I get one? The shepherd said yes. Thereupon the driver occupied himself in his vehicle for a good while and finally got out: "You have exactly 1658 sheep." The shepherd said, "That's right! Pick a sheep." The shepherd looked at him and finally said: "If I guess your profession, will you give me back the sheep?" "Sure, why not", the driver replies. "You are a biochemist", said the shepherd. The car driver was surprised: "That's right. How do you know that?" "Very simple," said the shepherd, "firstly, you come here although nobody has called you, secondly, you want a sheep as payment for telling me something I already know anyway, and thirdly, you have no idea what I do. Now please give me my dog back!"

And now for something completely different... ;-)


Lamarck said...

Hi (2),

and now for something completely different... ;-)

1. Biology is partially descriptive science and "lives" on precise definitions. When you talk about junk DNA, what exactly are you talking about? And what are your opponents talking about?

2. "Function", on the other hand, is an anthropocentric term (cf. the design principle "form follows function"). Function is by no means inferred from structure. For example: What is the function of a shoe with which I hammer a nail? To what extent is the analysis of the composition of a tyre relevant to understanding the function or essence of a car?

3. Since evolution is based on a "function follows form", potentially [sic!] functionless structures can be excluded. However, it could easily be hypothesized that any genetic structures - whether they encode bsw. proteins or not - simply act as a non-specific firewall. For example, as protection against ionizing radiation for genetic structures that can then be called conserved? As protection against generally thermodynamically caused, but then as lethal point mutations?

4. The individual organism is the carrier of the germ line in space and time. What matters is not the direct nature of the differences in the organismic construction of a species, but rather fault tolerance. The only condition here is to be able to carry the torch further. So it is easy for me to do without half a litre of blood, but may I call a corresponding yield in blood donation "junk blood"?

5. The debate essentially lacks the classification of structures that rely on fault tolerance. Mutation events occur over the entire life span - including, of course, on somatic cells - without the organism having to be significantly affected by them. Can the difference in sequence over time nevertheless be described as "junk" if it includes ageing processes - e.g. in connection with the onset of the menopause?

6. Viral fragments in the genome increase the carrier's resilience to new diseases. Does this affect the definition of the term?

7. Plesiomorphic fishes such as bichirs or lungfish have genome sizes of up to 139,000 Mbp. They do not have a gonosome, but have paired, asymmetric lungs and juvenile outer gills such as salamanders. In contrast, the genome size of modern puffer fish is 500 Mbp (Takifugu rubripes, at 365 Mbp, is the smallest known genome of a vertebrate). In Poeciliidae, Xiphophorus maculatus has an XY chromosome system, while Xiphophorus hellerii has a ZW system. In cross-breeding experiments with these two species, a completely new sex chromosome was created over about 100 generations. Doesn't a term such as junk DNA show an overly static view?

8. Evolution would not be possible if genetic structures did not show an extremely high form of redundancy. One form of redundancy is the endowment of almost every human cell with a complete genome and the necessary process control in a tissue network. Does your point of view have a wider scope than the inclusion of processed pseudogenes created by retrotransposition, although the possibility of such processes must first be proven here?



John Harshman said...

1. Junk DNA is sequence without sequence-related function and probably without bulk function.

2. One definition of function is "subject to selection". Another is "does something advantageous to the organism". While there may be ambiguous cases, in the main there seems no problem.

3., 4, 5 assume that mutations happen at a constant rate per genome, and that junk DNA attracts mutations away from functional sequences. This does not appear to be the case. Mutations happen in proportion to genome size.

6. I would be interested in documentation of that claim.

7. You point to the c-value paradox as if it supports some claim of yours. But it isn't sure what. No, there is not an overly static view.

8. Not sure why we should consider complete genomes in somatic cells as adaptive rather than simply the inevitable product of mitosis. Also not sure why you doubt the existence of and/or mechanism for processed pseudogenes.

Haven't you been dead for 200 years?

Lamarck said...


(a1) A sequence without a sequence-related function?! Here is some philosophy of science: things are recognized by their properties, but properties are not recognized by things. For example, nobody will define adenine by the fact that A has the function to form an A-T base pair with T. Because A can nevertheless be detected as such. Not to mention the fact that an A-U base pair can also be detected. So: Adenine is one of 4 nucleic bases... [Systematics], consisting of a purine backbone... [Structure], (possibly a little bit) [Logic, Evidence and Reason]. Define "junk DNA" without resorting to the notion of function.

(a2) What about an organism is not subject to selection?

(a3-5) With this I wanted to point to something else, but I would like to take up this: "Mutations happen in proportion to genome size." This would then be a detectable statistical influence that could be causally thermodynamically justified. Why should the conserved (= "older") regions of the genome have better repair systems than the easily mutable sequence regions? And if these easily mutable sequence regions are generally referred to as junk DNA, is it not the case that the genome as a whole has a regulatory influence on this amount of junk DNA?

(a6) DOI: 10.1371/journal.pgen.1000849; 10.1038/s41586-020-2486-3; 10.1126/science.aad5497.

(a7) If gonosomes are dynamic enough to emerge from non-gonosomes and then - as here - differentiate into an XY chromosome system as well as into a ZW system in closely related species, which can also be brought into an intermediate state by crossing: Where is junk DNA here?

(a8i) As part of the organism, somatic cells naturally have an indirect effect on the germ line and are therefore adaptive, regardless of whether there is relative susceptibility to cancer, infections, parasites, hereditary diseases or being eaten by a crocodile: it is difficult to live without intact somatic cells... Forensically, identical twins can also be distinguished on the basis of somatic cells. Are not the diverging mutation events in the somatic cells of identical twins an indication of junk DNA that can be regarded as relatively responsive?

(a8ii) Please show the existence of processed pseudogens.

("Haven't you been dead for 200 years?") I'm still alive, after all I am the inventor of the term biology.



João said...

Larry, have you ever read the essay "Betting on Chance - And no Fair Peeking" by Stephen Jay Gould? It deals with drift, selection, neutral theory, non-fucntional elements, and constraints. It's essay 27 in Eight Little Piggies.

João said...

*Essay 28

Larry Moran said...

It's an excellent essay.