More Recent Comments

Sunday, March 20, 2016

Another failure: "The Mysterious World of the Human Genome"

The Mysterious World of the Human Genome
by Frank Ryan
William Collins, an imprint of Harper Collins, London UK (2015)
ISBN 978-0-00-754906-1

This is just another "gosh, gee whiz" book on the amazing and revolutionary (not!) discoveries about the human genome. The title tells you what to expect: The Mysterious World of the Human Genome.

The author is Frank P. Ryan, a physician who was employed as an "Honorary Senior Lecturer" in the Department of Medical Education at the University of Sheffield (UK). He's a member of The Third Way group. You can read more about him at their website: Frank P. Ryan.

The first seven chapters are about the history of DNA. Ryan doesn't really tell us anything about what's in your genome until chapter 8. According to Ryan, the publication of the draft human genome sequences in 2001 led to several breakthroughs. Here's how he puts it on page 128.
The word 'breakthrough' is often misused in relation to scientific discovery, but here, indeed, was real breakthrough after breakthrough. And the breakthroughs presented a very much unprepared world of science with not one but three major surprises, each a challenging new mystery.
Regular readers of Sandwalk can probably guess what the three breakthoughs are because we've heard about them time and time again from amateurs who write about the human genome. Here's the list.
  1. Biochemists had figured out that there are about 100,000 proteins in human cells so they expected about the same number of genes. "It was an almighty shock" to discover that there are only 20,500 genes.
  2. The second major shock was that viruses or virus-like entities make up 45% of the genome.
  3. The third breakthrough is that 50% of the genome "appeared to code for nothing that we recognized at the time."
Frank Ryan suffers from a familiar syndrome called the The Deflated Ego Probblem. He describes his problem on page 130 ...
The paltry 20,500 genes seemed downright humiliating. To put it into perspective, we had roughly ten times as many genes as an average bacterium, four times as many as a fruit fly, and just twice as many as a nematode worm. In term of genes, we seemed hardly more complex than these humble life forms. A related revelation was the number of genes we have in common with these simpler organisms. We now discovered that we share 2,758 of our genes with the fruit fly and 2,031 with the nematode worm: and the three of us—human fly and worm—have 1,523 genes in common.
The truth is that the number of genes was pretty much what was expected by the experts [False History and the Number of Genes 2010] [Facts and Myths Concerning the Historical Estimates of the Number of Genes in the Human Genome]. And the truth is that "only" 1500 shared genes seems very low.

But if you were one of those scientists—like Frank Ryan—who didn't read the scientific literature, then you may have been under the false impression that humans need many more genes to make them feel important. In that case, your deflated ego needed puffing up so you had to come up with one or more of the seven possibilities that scientists use to rationalize the "shocking truth."

In Ryan's case, he tells us that the explanation was alternative splicing.
... understanding of how exons and introns work now affords an explanation of how just 20,500 genes could possibly code for 80,000 to 100,000 proteins.

A gene which, for example, had 14 exons separated by 13 introns, is likely to code for more than one protein. All that is necessary is that the regulatory mechanisms, which decide on which exons to splice together to make the messenger RNA, choose different combinations of exons. We now know that this is exactly what happens. The ability of a single gene to code for more than one protein is known as 'alternative splicing.'
(p. 135)
There are two things wrong with this description. First, human cells do NOT contain 80,000 - 100,000 different proteins. Second, alternative splicing is a real phenomenon but it only applies to a small number of genes. It is simply not true that most, or even many, human genes are spliced to produce different proteins [see Alternative Splicing]. Most examples of different forms of mRNA are just splicing errors that are never translated.

The number of virus sequences in the human genome is not a problem according to Frank Ryan. He tells us that several new genes have evolved from virus sequences and there may be many more.
In 2001, when the first draft of the complete human genome showed that roughly 45 per cent of the human genome appeared to be made up of retroviruses, or virus-like entities, such as LINES and SINEs, some biologists dismissed this huge genetic inheritance as junk, the graveyard of past viral infections. But today we have become a good deal more cautious in our interptetations.
The truth is quite different. The experts expected that most of our genome would turn out to be junk and that's exactly what was revealed when the human genome sequence was published. Half of our genome is composed of bits and pieces of ancient transposons. They look like junk and detritus and that's exactly what they are. Some small bits have secondarily evolved a function but the vast majority is still junk.

What about the rest of the genome? That's the subject of the third "breakthrough." Frank Ryan tells us that most of this DNA is required to regulate the expression of the 20,500 protein-coding genes. Some of that regulation involves epigenetics. Some of it contains stretches of DNA for ribosomal RNAs and tRNAs and some of it is devoted to a "surprsing variety of RNA molecules that did not code for proteins, but nevertheless had important roles to play in the control and expression of genes" (p. 188). There's a special class of these RNAs ...
... there is another more astonishing class of non-coding RNAs that regulates the human genome, a relatively new discovery that explains that mysterious black hole in the 2001 draft genome—the 50 per cent of our human DNA that was left a baffling blank. (p. 189)
He's talking about long non-coding RNA or lncRNA. After dutifully covering the few well-known examples of functional lncRNAs, Ryan continues with ...
Inspired by these discoveries, scientists began to search for more of these long non-coding RNA molecules to discover that they are transcribed pervasively throughout mammalian genomes. In time, lncRNAs were duly recognised as part of a newly recognised and very powerful epigenetic regulatory system, giving rise to an explosion of new research. This exciting new venture is still taking place as I write, but already we know that our human genome, like that of all plants and animals, contains vast number of long and small non-coding RNAs within which the lncRNAs comprise a class of their own, ranging in size from 200 to more than 100,000 nucleotides long. (p. 192)
You won't be surprised to discover that Frank Ryan repeat the myth that a gene was defined as protein-coding and you won't be surprised to learn that the recent discovery of genes for functional RNAs was a revolutionary finding that changed the way we look at genes.

The truth is far different. We've known about genes for functional RNAs since the 1960s and knowledgeable scientists since that time have never claimed that all genes have to encode proteins. We've known about a variety of small functional RNAs since the late 1970s—Sidney Altman got a Nobel Prize for one of them in 1989.

More of these small functional RNAs were discovered in the 1990s (a small surprise) but no knowledgeable scientist can defend the claim that most of our genome is devoted to their production.

The book closes with several chapters on the evolution of humans and the human genome. This is where Frank Ryan promotes his view of evolutionary change. His view includes mutation but also three other important mechanisms of change; epigenetics, symbiosis, and hybridization (of different species) (= MESH).

Epigenetics is trivial and unimportant in evolution. Symbiosis happened only a few times in three billion years. Hybridizations are one-off events that occur only once in every 10-100 million years, if at all. Mutations happen all the time.

Frank Ryan's goal is laudable ...
I set out to write this book from the premise that it would attempt to provide a non-scientific reader with a basic understanding of how his or her own genome works. I can only hope that I have succeeded in that aim. The very notion that we might understand the evolution, structural make-up and detailed function of the genomes that code for life, including our own human genome, is of epochal importance not only for science but also for all of us. I hope it has become clear that such understanding is important, since it must be for society in general, and not scientists alone, to decide where we go from here. (p. 298)
He has not succeeded. I don't say this just because I disagree with him. I'm certain that most of our genome is junk and unlike him I don't think the ENCODE Consortium was correct. That's not the issue. Frank Ryan is free to promote the minority view that most of our genome is functional but as a scientist writing for the general public he is obliged—in my opinion—to present the facts as fairly and objectively as possible. He has not done this. He doesn't present any of the evidence for junk DNA and he doesn't reveal to his readers any of the scientific objections that weaken his position.

Consequently, he leaves his readers with the impression that scientists have uncovered lots and lots of function in our mysterious genome. That means his non-scientific readers do not understand the real science behind the human genome after reading his book.

It's another failure, just like Nessa Carey's book Junk DNA: Why there is more to the human genome than meets the eye and John Parrington's book The Deeper Genome: A Journey Through the Dark Matter of the Genome. This is not how good science writers are supposed to behave.


Unknown said...

Symbiosis happened only a few times in three billion years.

You mean endosymbiosis here, right? Or actually you mean endosymbiosis where the endosymbionts end up as organelles of the host. Because there are certainly more examples of symbiosis and even obligatory endosymbiosis around.

Larry Moran said...

The author is only referring to symbiosis events that affect the evolution of the genome. Of course he covers endosymbiosis and the importance of mitochondria and chloroplast but he also refers to other examples such as the association of algae and fungi to form lichens and the association of rhizobial bacteria with legumes. In the latter examples, the emphasis is on the genetic changes in each of the organisms that make up the symbiosis.

He refers to a third kind of symbiosis such as the adaptation of hummingbirds to feed on nectar. I dismiss those examples as trivial—they are not substantially different from any other kind of adaptation to the environment.

If we take only the serious examples of symbiosis, including but not confined to endosymbioses, then this is hardly a ubiquitous phenomenon nor one that has occurred repeatedly over the history of life affecting a large number of genes.

Your quibble is noted ... perhaps I should have said about once every 100 million years and usually confined to a small clade. The message I'm trying to convey is that there's a huge, huge, difference in both time and extent between mutation and symbiosis as important creators of diversity. They don't belong in the same league.

Read the book if you want to continue the discussion. He also talks about transposons and viruses as examples of symbiosis but that's really just semantics.

Unknown said...

They don't belong in the same league.

It don't think they belong to the same category. I would have accepted gene transfer between symbionts as a genuine mechanism on a molecular level, but in all other cases we are talking just about adaptations and in some instances drift.

nmanning said...

"The author is Frank P. Ryan, a physician..."

No need to elaborate...

Matt G said...

I just watched "What Darwin Never Knew" again, and the errors surrounding junk DNA and the number of human genes jumped out at me. It's everywhere! Sean Carroll should know better.

Larry Moran said...

In humans, and many other complex species, genes occupy only a small fraction of all of the DNA, and are separated by long intervals of noncoding DNA. Some of this noncoding DNA functions in the control of how genes are used, but a lot of it is what is called "junk." This junk accumulates by various mechanisms and often contains long repetitive tracts with no information content. I will generally ignore this junk, but it is worth mentioning in order to have a picture of the structure of our genomes as archipelagoes of islands (genes) separated by vast areas of open sea (junk.) Sean B. Caroll, "The Making of the Fittest" p. 76)

Most dark matter contains no instructions and is just space-filling "junk" accumulated over the course of evolution. In humans, only about 2 to 3 percent of our dark matter contains genetic switches that control how genes are used. Sean B. Carroll, "Endless Forms Most Beautiful" p. 112

Matt G said...

I'm glad he knows it. He should have looked over the video based on his book, and featuring him prominently, before it went out. More people will watch the video than read the book.

Unknown said...

Dear Larry,
Epigenetics is not trivial when there is epigenetic inheritance over several generations.
Hybridizations occur frequently in wild plants. Wheat, potato, tomato, cabbages, etc. are hybrids.
New symbiosis have been observed in the lab (Pak, J.W., and Jeon, K.W. (1997). J. Eukar. Microbiol. 44, 614-619)

Larry Moran said...

Epigenetics is trivial. We've known for decades that bacteria inherit methylation at restriction sites and we've known for decades that the progeny of bacteria growing in glucose in the absence of lactose inherit lac repressor and a silenced lac operon.

We've known for decades that yeast daughter cells inherit silenced regions of the genome at the Hidden MAT loci (HML or HMR) and we've known for decades that constitutive heterochromatin regions at the centromeres of eukaryotic cells is heritable.

Not of this has caused a significant stir among evolutionary biologists because it's trivial on the time scale of evolution. The best thing that can be said about epigenetics is that in a very few rare instances it may actually lead to real genetic changes in the genome but this is an indirect effect that's no different from a lot of other things that may potentiate certain mutations.

"Epigenetics" is biology's version of the gluten-free phenomenon. There's a little bit of truth in there somewhere but most of it is hype and misunderstanding promoted by people who epitomize the saying that "a little knowledge is a dangerous thing."

Larry Moran said...

Hybridizations do occur, of course. There are, for example, about 350,000 species of flowering plants with most lineages stretching back about 200 million years and quite a few hybridization events have occurred in those lineages. I don't know how many but maybe it's as high as 100,000. That's being generous and only counting hybridization events between genuine species that have been reproductively isolated for a long time.

So, there would have been one of these events every 20 years somewhere among the many branches of a large tree—mostly between adjacent twigs. In contrast, every new generation of plants has hundreds of new mutations. These events aren't in the same ball park.

Frank Ryan suggests that the mating of humans and Neanderthals is an example of hybridization but that's only significant if these are two different species. Otherwise it's no more significant than mating between subpopulations (races) that are part of the same species.

(BTW, I don't count artificial selection as part of natural biological evolution.)

The issue here is not whether epigenetics, hybridization, and symbiosis occur, it's whether their relative frequency is sufficient to count them as common contributors to change in the frequencies of alleles in a population. In other words, should they be incorporated into standard evolutionary theory or should each event be treated as one-off examples in the history of life?

Anonymous said...

" Recent developments in genomics are revolutionizing our views of angiosperm genomes, demonstrating that perhaps all angiosperms have likely undergone at least one round of polyploidization and that hybridization has been an important force in generating angiosperm species diversity." -- Soltis & Soltis 2009.

Of course, the authors are sweeping the hybridization issue under the polyploidy umbrella, and I can't get further information at the moment without buying the article.

I'm not trying to say hybridization is a more important source of variation in plants than is mutation within lineages. Nonetheless, it's big.

One can quibble about where the hybridization occurs in relationship to speciation itself, but I think two things are important here.

First, it seems to me that selection against hybridization can occur only after speciation has happened and the resulting offspring are of low fitness. Of course, inability to hybridize can develop accidentally, too, and can sometimes actually be the cause of the speciation.

Second, I think that in plants there may not be the same selection pressure for preventing hybridization as there is in vertebrates. Most individual plants produce a lot more offspring than individual vertebrates, and the seeds produced by one maternal plants have several different fathers. So "wasting" a few as hybrids may have low cost, and sometimes those hybrids will pay off. Plants can't move to seek a better environment; offspring might benefit from having a set of genes that would have been harmful in the parent in its environment.

Some plants seem set up for serious hybridization. The variable biological unit we call Kentucky Bluegrass (Poa pratensis) for example, consists of plants that share maternal genetic material but gain additional variation through hybridization; the offspring of hybridization join the interbreeding Poa pratensis unit. Poa pratensis also sets seed apomictically, without sex.

Anyway. I should go have breakfast and stop rambling.

John Harshman said...

Big deal. All vertebrates have also undergone at least two rounds of polyploidization, and teleosts had undergone three. Note that Soltis & Soltis are referring to a polyploidization event at some unspecified time in the past. Allopolyploidy may be responsible for as much as 5% of speciation events in angiosperms according to some estimates, but that's still a fairly small percentage of the total.

Anonymous said...

True. Hybridization is so important in some plants I work with that my view can become biased.

txpiper said...

Speaking of another not-so-candid admission of the inadequacy of the theory, an exciting new concept emerges.

"We now know that intuition fails us, with feathers, eyes and all living things the product of an entirely natural process."

Of course, we actually don't now know any such thing. Highly educated minds (never underestimate the power of relentless inculcation) consciously piss themselves away, but in the dark and naked recesses of mutant minds, the doubts persist. And they damn well should.

Fair Witness said...

"..mutant minds...". From an intelligent designer? Blasphemy !

Unknown said...

Not every reader of a genomics/ genetics/ biology book is a scientist or academician.

Some are plain joes/ janes who are just out to satisfy their curiosity of the world. Of these laypeople, some haven't touched a book on science for 20 years or more. I am one of the latter group.

Sure, going by your scathing review, there seem to be generalizations and oversimplifications in this book, but in order to appeal to the reader like me, scientific accuracy and exactitude may not always work. For with accuracy and exactitude come technical jargon too. And jargon often puts off the lay reader.

I hope you get what I'm saying. I guess the author has taken some liberties in order to appeal to a more general audience.

And as far as the 'epigentics' topic is concerned, what I've seen from my readings, albeit general in nature, is that there are clearly two sides of the fence: those for epigeenetics and those staunchly against. I guess you might belong to the latter. But I'm not going to dispute you. Not having the requisite technical wherewithal to lean on, there's no way I'm going to ever know who is right, those for or those against.

Now however, Amazon tells me you're the author of a university-level textbook on biochemistry, which has run into several editions right from 1994. So it seems you too would be a good author whose book(s) I should read. I shall thus wait for the day you bring out a popular science treatise.

I hope I haven't bored you

Larry Moran said...

I assume you've read the book from cover to cover.

Did you learn from reading the book that there are "clearly two sides of the fence" on the subject of epigenetics? If not, do you think you were deceived and mislead because you weren't told about the controversial nature of Frank Ryan's view?

Do you think this is the proper way to explain science to a layperson?