Like most pharmaceutical companies, Amgen produces promotional videos that are supposed to show how close they are to cutting edge science. Here's one that celebrates DNA Day and answers the question "What Is the Dark Genome? 98% of DNA explained." Are you impressed?
What Is the Dark Genome?
21 comments :
Why would I be impressed when the first thing I see after clicking the link is a fragment of left-handed DNA
I'm guessing they haven't read your book.
I'm impressed with how they manage to hire only people who do not understand molecular evolution.
I still find it difficult to understand why scientists don't understand basic principles such as the composition of genomes. Don't they ever talk to each other about concepts such as junk DNA, molecular evolution, and regulation?
It would be interesting to know how many molecular biologists have taken any courses in evolutionary biology or molecular evolution.
There were colleagues of mine in my department who quietly told their graduate students that they did not need to take my population genetics course. They were molecular biologists -- and very good ones in other respects.
There’s now a Casey Luskin post at Science and Culture Today crowing about an article in a prestigious journal which announces a Paradigm Shift to declaring transposons in the genome to largely be functional. It has something like 258 citations. The journal is Cell.
That “Anonymous” was me, Joe Felsenstein.6
How/Why did Cell publish that?
Obviously because it's stimulating, controversial, and because all true molecular biologists and genomicists know it's got to be true -- except for a small percentage of them who are evolutionary types.
Or perhaps it's typical science hype: your work is so much more important and tenure-worthy if it revolutionizes our understanding of the world.
Back about 1980, when everyone started saying they were introducing a New Paradigm, I decided that I would become unique among scientists of my generation by being the only one doing Normal Science.
@John Harshman
In my opinion, there's a difference between hype and lies. This Cell paper crosses the line and misrepresents the facts and that's lying. I think the Amgen scientists are also misrepresenting the facts.
Two points: It's not lying if you believe it. And it's not misrepresenting the facts, exactly; it's drawing unwarranted conclusions from a biased sample, e.g. finding a few functional TEs and claiming that therefore all TEs are functional.
I agree that they aren't lying, and the fact that they aren't lying is shameful to the scientific community. Imagine if exoplanet researchers declared that every planet in the universe is habitable just because we found some planets with water on them. That's where we are with the disaster of adaptationism in biology. Personally, my theory is that it's a combination of adaptationism and fetishization of complexity. Biologists are so obsessed with declaring how complex everything is that they assume that just because something happens in a cell it must be part of some mysterious super important unknown pathway that ignorant biologists of the past would have just called "junk". Even if you do a knockout experiment and show it doesn't matter they'll try to rationalize it as being used for some special cell type, life stage, or environment, which is fair enough as a hypothesis, but they take the mere possibility of a function as proof of a function to the point that it's unfalsifiable.
I’m starting to think that you and Joe should follow up on every example of this you see coming from scientists and write them a long referenced explanation on why they should be cautious in their claims. Over time this effort might have an impact
The current edition of Alberts is cautious on the 80% functional claim. I was a reviewer for the previous edition of Lodish and suggested they pull back some of their previous claims. The next edition will be 1-2 years so we’ll see
@Lantog
I help a bit on the last (7th) edition of Alberts. (Bruce Alberts was my Ph.D. supervisor.) Here's what they say on page 230.
"The first obstacle in interpreting the sequence of the 3.1 billion nucleotide pairs in the human genome is the fact that the nucleotide sequence of most of it (about 90%) is probably functionally unimportant."
"... The important question of how much of the DNA sequence of the human genome is functionally important was briefly confused by a set of high-profile publications that appeared in 2012 from a large, federally funded US genome project named ENCODE. These publications, which reported the results of a massive survey using sensitive assays that can detect the presence of RNA molecules in cells at extremely low levels, reported that 76% of the total DNA sequence in human cells is transcribed to produce RNA molecules. Even though many of these transcripts were found at levels of less than a single RNA molecule per cell, the ENCODE scientists used such data to assert that most of human DNA is functional, with very little "junk." This claim received widespread publicity, along with their belief that our genome contains tens of thousands of previously undetected genes that produce RNA molecules that do not code for protein."
continuing the quote from Alberts et al. 7th ed. (p. 231)
"As previously stated, there is a strong scientific consensus that most of the human genome consists of DNA whose nucleotide sequence is not relevant to biological function - being so-called junk. This conclusion rests on the finding that natural selection fails to preserve these sequences in the face of the inevitable random changes to genomes that occur over time, as can be seen both when different species are compared and from detailed analyses of human variation. The fact that these DNA sequences nevertheless produce an occasional RNA molecule can be explained by the occurrence of background "noise" in gene expression. Although gene expression is very accurate, it is not perfect, and biochemical errors occasionally occur. Such errors are to be expected, and so long as they are kept at a low level, they are thought to have little or no consequence for the cell."
Wouldn't it be nice if research scientists, postdocs, and graduate student read a reputable textbook from time to time?
I wonder how many students actually use textbooks today. My university has installed some AI tool in which they can feed pdfs of our lectures. It sumarizes them and develops questions they can use for learning. I don't know if this really works but I guess there is no way back.
Post a Comment