More Recent Comments

Tuesday, June 27, 2023

Gert Korthof reviews my book

Gert Korthof thinks that the current view of evolution is incomplete and he's looking for a better explanation. He just finished reading my book so he wrote a review on his blog.

Scientists say: 90% of your genome is junk. Have a nice day! Biochemist Laurence Moran defends junk DNA theory

The good news is that I've succeeded in making Gert Korthof think more seriously about junk DNA and random genetic drift. The bad news is that I seem to have given him the impression that natural selection is not an important part of evolution. Furthermore, he insists that "evolution needs both mutation and natural selection" because he doesn't like the idea that random genetic drift may be the most common mechanism of evolution. He thinks that statement only applies at the molecular level. But "evolution" doesn't just refer to adaptation at the level of organisms. It's just not true that all examples of evolution must involve natural selection.

I think I've failed to explain the null hypothesis correctly because Korthof writes,

It's clear this is a polemical book. It is a very forceful criticism of ENCODE and everyone who uncritically accepts and spreads their views including Nature and Science. I agree that this criticism is necessary. However, there is a downside. Moran writes that the ENCODE research goals of documenting all transcripts in the human genome was a waste of money. Only a relatively small group of transcripts have a proven biological function ("only 1000 lncRNAs out of 60,000 were conserved in mammals"; "the number with a proven function is less than 500 in humans"; "The correct null hypothesis is that these long noncoding RNAs are examples of noisy transcription", or junk RNA"). Furthermore, Moran also thinks it is a waste of time and money to identify the functions of the thousands of transcripts that have been found because he knows its all junk. I disagree. The null hypothesis is an hypothesis, not a fact. One cannot assume it is true. That would be the 'null dogma'.

That's a pretty serious misunderstanding of what I meant to say. I think it was a worthwhile effort to document the number of transcripts in various cell types and all the potential regulatory sequences. What I objected to was the assumption by ENCODE researchers that these transcripts and sites were functional simply because they exist. The null hypothesis is no function and scientists must provide evidence of function in order to refute the null hypothesis.

I think it would be a very good idea to stop further genomic surveys and start identifying which transcripts and putative regulatory elements are actually functional. I'd love to know the answer to that very important question. However, I recognize that it will be expensive and time consuming to investigate every transcript and every putative regulatory element. I don't think any lab is going to assign random transcripts and random transcription factor binding sites to graduate students and postdocs because I suspect that most of those sequences aren't going to have a function. If I were giving out grant money I give it to some other lab. In that sense, I believe that it would be a waste of time and money to search for the function of tens of thousands of transcripts and over one million transcription factor binding sites.

That not dogmatic. It's common sense. Most of those transcripts and binding sites are not conserved and not under purifying election. That's pretty good evidence that they aren't functional, especially if you believe in the importance of natural selection.

There's lot more to his review including some interesting appendices. I recommend that you read it carefully to see a different perspective than the one I adocate in my book.


Mikkel Rumraket Rasmussen said...

Everyone can fall victim to dichotomous thinking. To say that the fixation of most mutations owes to genetic drift isn't the same as saying natural selection is not an important contributing cause in evolution.

And then many people also have issues with the idea of multiple causation. We like to ask "what is the cause of X", but in reality we more often find that "X has been influenced by the factors A, B, C, etc."

John Harshman said...

A study wouldn't be all that expensive if you think in statistical terms. Just randomly sample the supposed functional elements and infer tha the frequency of function in unsampled elements resembles that in the sampled ones. Of course for that purpose it's crucial that the sample be unbiased. But that shouldn't be hard.

Joe Felsenstein said...

The bad news is that I seem to have given him the impression that natural selection is not an important part of evolution.
Told you so!

SPARC said...

You should not miss Alex Palazzo's discussion of Geert's review. You will find it on his blog under The importance of charity in making forceful arguments and Determinants of genome size

Larry Moran said...


I'd like to draw your attention to the comment I made on Alex's blog post three weeks ago and his response. It will help clarify his intention.

Adaptationsts get very upset whenever we emphasize drift as an important mechanism of evolution. They often accuse us of ignoring natural selection but that's not true. The real problem is them and their constant refusal to recognize the role that random genetic drift plays in evolution. Pot kettle black.

SPARC said...

Thanks, I should have scrolled further down when I read Alex post.