Paradigm shifting

I was reading up on non-coding RNAs and came across this recent paper.

Bhartiya, D., and Scaria, V. (2016) Genomic variations in non-coding RNAs: Structure, function and regulation. Genomics 107:59-68. [doi: 10.1016/j.ygeno.2016.01.005]

Abstract: The last decade has seen tremendous improvements in the understanding of human variations and their association with human traits and diseases. The availability of high-resolution map of the human transcriptome and the discovery of a large number of non-protein coding RNA genes has created a paradigm shift in the understanding of functional variations in non-coding RNAs. Several groups in recent years have reported functional variations and trait or disease associated variations mapping to non-coding RNAs including microRNAs, small nucleolar RNAs and long non-coding RNAs. The understanding of the functional consequences of variations in non-coding RNAs has been largely restricted by the limitations in understanding the functionalities of the non-coding RNAs. In this short review, we outline the current state-of-the-art of the field with emphasis on providing a conceptual outline as on how variations could modulate changes in the sequence, structure, and thereby the functionality of non-coding RNAs.

The concept of a scientific paradigm and a "paradigm shift" was promoted by Thomas Kuhn. His most popular work was The Structure of Scientific Revolutions, published in 1962.

One of the ideas expressed in that book is that scientists operate on a day-to-day basis with a common set of assumptions about their discipline. Those "assumptions" are like a model—they are supported by a large amount of data and/or theoretical studies so that the average scientist in the field doesn't question the model. Most of the work in the discipline is devoted to discoveries within the "paradigm" and not to overthrowing it.

Gradually, evidence may accumulate that questions or challenges the existing model. Usually that data is set aside and counted as anomalous, an exception, or an artifact. In other words, the model resists change in contrast to the "nasty, ugly, little fact"¹ view of science mentioned by T.H. Huxley.

This is a simplistic view of Kuhn's thesis, to be sure, but I'm going to talk about "paradigm shifting" in the colloquial sense, so bear with me.

This is a good time to note that many (most?) philosophers of science disagree with Kuhn's view of how scientists operate. I think that's largely because Kuhn was writing about the sociology of the scientific community and not epistemology. No doubt some philosophers will weigh in. So far, they have not been very effective in preventing the abuse of the word "paradigm" whether they believe in it or not.

I sympathize with Stephen Jay Gould who wrote about this in "A Foot Soldier for Evolution," one of the essays in the anthology Eight Little Piggies. Gould was a friend of Thomas Kuhn and he agreed with the general thrust of the argument in The Structure of Scientific Revolutions. In the following excerpt he's referring to the idea that a nasty, ugly, fact could kill a beautiful theory. Gould says,

Some beliefs may be subject to such instant, brutal and unambiguous rejection. For example: no left-coiling periwinkle has ever been found among millions of snails examined. If I happen to find one during my walk on Nobska beach tomorrow morning, a century of well nurtured negative evidence will collapse in an instant.

This Huxleyan vision of clean refutation buttresses one of our worst stereotypes about science. We tend to view science as a truth-seeking machine, driven by two forces that winnow error: the new discovery and the crucial experiment—prime generators of those nasty,ugly, little facts. Science does, of course, seek truth, and even succeeds reasonably often, so far as we can tell. But science, like all of life, is filled with rich and complex ambiguity, The path to truth is rarely straight, marked by a gate of entry that sorts applicants by such relatively simple criteria as age and height. (When I was a kid, you could get into Yankee Stadium for half price if your head didn't reach a line prominently drawn on the entrance gate about four feet above the ground. You could scrunch down, but they checked. One nasty, ugly, day, I started to pay full price, and that was that.)

Little facts rarely undo big theories all by themselves—the myth of David and Goliath notwithstanding. Such facts can refute little, highly specific theories, like my conjecture about lefty periwinkles, but they rarely slay grand and comprehensive views of nature. No single, pristine fact taught us that the earth revolves around the sun or that evolution produced similarities among organisms. Overarching theories are much bigger than single facts, ...

Instead, little facts are assimilated into large theories. They may reside there uncomfortably, bothering the honorable proponents. Large numbers of little facts may eventually combine with other social and intellectual forces to topple a grand theory.

What did Bhartiya and Scaria mean when they mentioned "paradigm shift" in their abstract? What "paradigm" got shifted and what were the facts that caused such a shift?

We get some clues in the introduction where they write,

One of the major discoveries that have come up in the last decade, with comprehensive maps of genome diversity, transcriptome and epigenome at hand, was the discovery of pervasive transcription in the human genome.

The references they give are to two reviews with a common author, Alain Jacquier of the Pasteur Institute in Paris (France) (Jacquier, 2009; Jensen et al., 2013).

I'm assuming that the idea of pervasive transcription is new to Bhartiya & Scaria and I'm assuming they're fans of the hypothesis that most of these RNAs have a function. But just because some idea is new to you, doesn't mean that it conflicts with the prevailing model of the experts. So Bhartica & Scaria checked the literature and found two recent reviews published in high quality journals. Both of them emphasized the "revolutionary" idea of pervasive transcription and the likely functions of all those transcripts.

If it's true that large complex genomes have many more genes for functional RNAs than for proteins then this really is a big change in our understanding, especially if most of those RNAs are required for the sophisticated regulation of the protein-coding genes. This conflicts with the worldview (paradigm) of a messy genome full of junk and the view that our old understanding of gene regulation by transcription factors is basically correct.

If that's the case, then a paradigm shift is under way.

However, the idea of pervasive transcription is not new to everyone since it has been known for 45 years. If many of the real experts in the subject knew about pervasive transcription long before publication of the human genome sequence then where is the paradigm? I submit that the word "paradigm" should only be used to describe the dominant worldview among experts in the field and by "dominant" I mean overwhelming.

There are lots and lots of scientists (I am one) whose model of how biology works at the molecular level hasn't changed substantially in 30-35 years. They've known for a long time that there are many different kinds of genes including protein-coding genes and a variety of different genes for RNAs. Those genes include genes for regulatory RNAs (discovered in the late 19070s).

They've known for a long time that most of the genome is transcribed but the transcripts from intergenic regions tends to be of very low abundance. This knowledge dates from hybridization studies in the early 1970s. They know that promoters and transcription factor binding sites occur frequently in large genomes by accident so that there MUST be spurious binding as a consequence of the properties of DNA binding proteins. Where there's spurious binding there will be spurious transcription. It's a logical consequence of the properties of DNA and large genomes.

They've known from molecular studies of Drosophila, yeast, and C. elgans development that large phenotypic changes happen with small changes to regulatory sites that control expression of developmental genes. This is the contribution of evo-devo in the 1980s. It means that most animals, for example, will have pretty much the same complement of genes.

They've known that most of the DNA in large complex genomes is junk. They've known that this is consistent with modern evolutionary theory ever since the revival of random genetic drift and nearly-neutral theory over 45 years ago.

All of these ideas were confirmed by the publication of the human genome sequence. The only significant change in the past few decades—beginning about 1990—is that there are more genes for small RNAs than most of us expected but the differences aren't huge. There may be 5000 such genes in the human genome² wheres if you had asked me in 1990 I would have said there were only a few hundred. This change is easily accommodated into my worldview (paradigm) and those of other scientists.

It's true that there were many scientists who assumed an entirely different model. I'm convinced that critics who complain about the word "gene" are correct in one sense, most scientists thought and acted as though all genes encoded proteins. It's certainly true that the majority of textbook writers and teachers thought this way.

But if a large number of scientists and teachers believe in a model that's demonstrably false, it this a scientific "paradigm"? I don't think that's what Kuhn had in mind. If something happens to correct those false ideas, is that a "paradigm shift"? I don't think so although I'll readily admit that it's a revolution of some sort.

Let's summarize the argument so far—I apologize for rambling on about this.

1. Bhartiya & Scaria believe that the prevailing model in 2000 was that most (all?) genes encoded proteins and these genes made up only a small percentage of the genome. This was the prevailing paradigm, in their view.
2. Bhartiya & Scaria believe that pervasive transcription was only discovered about 10 years ago and this led to the realization that there are huge numbers of genes for non-coding RNAs covering almost all of the genome.
3. These new discoveries have overthrown the prevailing paradigm leading to a paradigm shift.

The logic here is correct but there's something deeply flawed with this view of genomics. Bhartiya & Scaria are wrong about the prevailing paradigm. They're wrong about the discovery of pervasive transcription. And they're wrong to claim that most of those transcripts are functional.

There is no paradigm shift and that phrase should not have been used in a review published in a reputable journal. The reviewers of the journal Genomics should have exercised better judgement.

I'm working my way through Suzan Mazur's book The Paradigm Shifters. It's a series of light-weight interviews with scientists who view themselves on the leading edge of a revolution in the way we think about evolution.

The funniest thing about this group is that they're all "shifting" in different directions! The saddest thing is that they're all fighting a "paradigm" that only exists in their own imaginations. It's a backward, outdated view of evolution that dates back to the nineteenth century. Most of them missed the real revolution that took place with the development of modern population genetics.

Like so much of the rhetoric about paradigm shifting, they are doubly wrong. They don't understand the "paradigm" and their "shifts" are either incorrect or insignificant.

Stop using the word "paradigm." It doesn't mean what you think it means. Stop claiming that a revolution is under way unless you can show that your facts are correct and the views of the experts are changing. Most challenges to the prevailing model are bound to fail. That's how the system works and that's a good thing.

Is it possible to recognize a real paradigm shift when it is under way? I think not, especially in the early stages where, by definition, scientists are skeptical of the new results.

Are there any good examples of "paradigm shifts" in biochemistry and molecular biology? I don't think there are very many. Chemiosmotic Theory is the best example I can think of [Nobel Laureate: Peter D. Mitchell]. Most other new discoveries were easily incorporated into existing models (e.g. introns and splicing).

---

1. "A beautiful theory, killed by a nasty, ugly, little fact." Thomas Henry Huxley speaking to Herbert Spencer as reported by Francis Galton.

2. My money is on less than 1000.

Jacquier, A. (2009) The complex eukaryotic transcriptome: unexpected pervasive transcription and novel small RNAs. Nature Reviews Genetics, 10:833-844. [doi: 10.1038/nrg2683]

Jensen, T.H., Jacquier, A., and Libri, D. (2013) Dealing with pervasive transcription. Molecular cell, 52:473-484. [doi: 10.1016/j.molcel.2013.10.032]