Wednesday, January 16, 2013

Why Do the IDiots Have So Much Trouble Understanding Introns?

Most eukaryotic genes have introns. Introns make up about 18% of the DNA sequences in our genome. Most of these sequences are junk but introns are functional and up to 80bp of each intron is required for proper splicing. The essential sequences contain the 5′ splice site (~10 bp); the 3′ splice site (~30 bp): the branch site (~10 bp); and enough additional RNA to form a loop (~30 bp). The branch site and the splice sites are where specific proteins bind to the mRNA precursor [Junk in Your Genome: Protein-Encoding Genes]. It turns out that within introns about 0.37% of the genome is essential and about 17% is junk.

The primary transcripts of some genes are alternatively spliced. In other words, there are at least two different versions of mRNA produced from the same primary transcript. We've been teaching alternative splicing for thirty years—it's been in the textbooks for decades. The inclusion or exclusion of an exon during splicing is controlled by regulatory proteins that bind to specific sites within the intron—often near or at the 5′ splice site or the 3′ splice site. The binding sites are, at most, 10bp in size. Even if there were five of them per intron this would only mean an additional 50bp of essential sequence and it would only apply to a small percentage of all introns. Not enough to worry about when calculating the total amount of junk DNA in our genome.

Casey Luskin worries about such things [Yet Another Blow to "Junk DNA": Paper Shows How Introns Are Key to the Splicing Code]. He found a recently published paper by Wang et al. (2013) that identified a handful of new splicing regulatory factors and binding sites. This leads Casey Luskin to conclude that ...
Studies like this one add to the weight of evidence showing introns are vital -- in this case, for splicing of RNA to create important protein-coding transcripts. The "junk" in the genome turns out, once again, to be very much otherwise.
It seems as though poor Casey has just learned for the first time that there are RNA binding sites in introns. It's apparent that he does not have the knowledge to calculate how much of an effect this might have on the total amount of junk DNA in our genome.


Genomes & Junk DNA

Alternative splicing has been a problem not only for Casey Luskin but also for Richard Sternberg, who actually has some training in biology (i.e. fewer excuses). A few years ago (June 2010) he published an article in Evolution News & Views (sic) where he claimed that 171,000 introns in the human genome participate in alternative splicing pathways. He believes (incorrectly) that 90% of human genes are alternatively spliced. There are about 190,000 introns in 20,500 protein-encoding genes for an average of about 9 introns per gene.² According to Sternberg, since 90% of 190,000 = 171,000 it means that there are a lot of "functional" introns. They can't be junk [Creationists, Introns, and Fairly Tales]. Sternberg made a serious error in his calculation—one that could not be made by anyone who understands alternative splicing.¹

When the error in his calculations was pointed out to him he repeated the mistake in another post [IDiots Do Arithmetic a Second Time - Same Result].

Jonathan McLatchie [Alternative Splicing and Why IDiots Don't Understand How Science Works] and Jonathan Wells [Jonathan Wells Weighs in on Alternative Splicing] have also demonstrated an inability to understand splicing and alternative splicing even though they write about it on creationist blogs.

This is why I gave a talk recently on Science vs IDiots where I pointed out that most creationists don't know enough about biochemistry, molecular biology, and evolution to pass an undergraduate course. Yet that doesn't prevent them from criticizing papers on all three topics. And no matter how often you point out their lack of knowledge, it doesn't seem to stop them from making the same mistakes over, and over, and over again. Isn't that strange?

1. Even if 90% of our genes are alternatively sliced—they aren't—that only means that at least one intron is involved in alternative splicing events. The best you can say is that 90% ×: 23,000 genes = 21,000 genes are alternatively spiced and as few as 21,000 introns participate in alternative splicing.

2. A better average is 7.2 introns per gene [Junk in Your Genome: Intron Size and Distribution].

Wang, Y., Xiao, X., Zhang, J., Choudhury, R., Robertson, A., Li, K., Ma, M., Burge, C.B., and Wang, Z. (2013) A complex network of factors with overlapping affinities represses splicing through intronic elements. Nature Structural & Molecular Biology 20:36-45. [doi:10.1038/nsmb.2459]


  1. It should also be noted that it is debatable how much alternative splicing is actually functionally important, vs. just unavoidable noise in an imperfect splicing process. The fact that many multicellular organisms with smaller genomes get by with much smaller introns and fewer introns indicates that probably a lot of them are not all that important.

    And: functional retrogenes.

  2. Great post. I read the thing about Sternberg from 2010. Wow... Just wow.

    As for Casey Luskin, you all (Larry included) must read this post at Sensuous Curmudgeon, which is about how the Discovery Institute used a bogus, biased poll of a few Louisiana teachers to promote the Louisiana anti- Science re- Education Act (LSEA).

    The poll was biased and based on tiny numbers, but to make it better, Casey Luskin misreported the figures. 55% of their tiny sample of teachers said they did NOT feel intimidated when teaching about "Origins."

    That didn't serve the IDiots' political purposes, so lawyer Luskin just changed the numbers: Casey said 55% of their tiny sample of teachers reported feeling they DID feel intimidated when teaching about "Origins."

    So the whole "Academic Freedom" whine, used to push through the LSEA and other laws in other states, was another hoax by ID creationists at the DI.

    Curm cites Prof. Barbara Forrest who gives the details about the Louisiana ID hoax.

    You'll recall Prof. Forrest terrified the DI when she was asked to testify as an expert witness at Dover: the DI called her an internet stalker and tried to block her testimony. They cry "academic freedom" but they wanted to silence Forrest. They cry "we're martyrs" and "teach the controversy" but they wanted to face ZERO opposition at Dover.

    Well, Forrest did testify at Dover, she revealed the search-and-replace blunder on the Michael Behe et al's textbook "Of Pandas and People" that replaced the word "creationist" with "intelligent design" immediately after the Supreme Court banned creationism in 1987, resulting in "cdesign propentsist."

  3. IMO the funny thing is that Casey refers to the methodology used in the paper which shouldn't work if his creationist ideas were true (he actually cites another blog):
    "These introns of the reporter gene carried random DNA sequences. When the reporter is screened and shows green it means that portion of the intron is spliced."
    This arguement is outright stupid. If he argues that these elements are designed the methodology shouldn't work and if he argues that they are too short to contain enough of what he may call CSI to prevent selection they cannot be designed.
    In addition, he (IMHO willfully) doesn't mention that
    "Many factors bound multiple distinct motifs with similar affinity, and all motifs were recognized by multiple factors, which revealed a complex overlapping network of protein-RNA interactions." which would better fit to an evolved and still evolving mechanism rather than to his picture of an in-depth designed fine-tuned cell.

  4. SPARC,
    I'm not sure I get it:

    This arguement is outright stupid. If he argues that these elements are designed the methodology shouldn't work and if he argues that they are too short to contain enough of what he may call CSI to prevent selection they cannot be designed.

  5. 1. Pedantry - in the context of this essay, genes are not spliced. RNAs are.

    2. Point of clarification - alternative splicing is controlled, not just by intronic elements, but also by sequences that reside within exons. Of course, this does not help the IDists in their arguments. After all, if so much splicing "function" resides in exons, then that's just less functionality within introns, and thus more junk in introns.

    1. 1. I fixed the pedant point.

      2. You are correct.

  6. Re: Casey's introns are vital assertion; I wonder how they will rationalize the paper below. It seems the that nature can delete the introns with no issue.

    Ciomborowska et al. (2013) “Orphan” Retrogenes in the Human GenomeMol Biol Evol 30: 384-396.


    Gene duplicates generated via retroposition were long thought to be pseudogenized and consequently decayed. However, a significant number of these genes escaped their evolutionary destiny and evolved into functional genes. Despite multiple studies, the number of functional retrogenes in human and other genomes remains unclear. We performed a comparative analysis of human, chicken, and worm genomes to identify “orphan” retrogenes, that is, retrogenes that have replaced their progenitors. We located 25 such candidates in the human genome. All of these genes were previously known, and the majority has been intensively studied. Despite this, they have never been recognized as retrogenes. Analysis revealed that the phenomenon of replacing parental genes with their retrocopies has been taking place over the entire span of animal evolution. This process was often species specific and contributed to interspecies differences. Surprisingly, these retrogenes, which should evolve in a more relaxed mode, are subject to a very strong purifying selection, which is, on average, two and a half times stronger than other human genes. Also, for retrogenes, they do not show a typical overall tendency for a testis-specific expression. Notably, seven of them are associated with human diseases. Recognizing them as “orphan” retrocopies, which have different regulatory machinery than their parents, is important for any disease studies in model organisms, especially when discoveries made in one species are transferred to humans.

    1. Also, for retrogenes, they do not show a typical overall tendency for a testis-specific expression.

      Are retrogenes supposed to be normally expressed in your balls?

    2. This comment has been removed by the author.

    3. It seems the testis have somewhat over-active expression. Something only expressed there is either really interesting for reproductive biology, or spurious expression. Hard to tell which.

      But is a pseudo-gene-looking thing is only expressed in the testis is normally not a sign of a biological function.

  7. Larry,

    what kind of real evidence do we have on how widespread functionally relevant alternative splicing is?

    You put it at maybe 5%. On what basis can we guess that?

    1. We have very few examples of biologically functional alternative splicing. On the other hand, we have plenty of evidence that most examples of proposed alternative splicing are probably splicing artifacts or errors.

      The high proposals (>90%) are just ridiculous—they make no biological sense. I'm suggesting 5% based on the known examples of alternative splicing, which seem to be quite rare. It's just an educated guess designed to provoke discussion.

      Here are some links from previous posts.

      A Challenge to Fans of Alternative Splicing

      Splicing Error Rate May Be Close to 1%

      Extraordinary Claims about Human Genes

      How to Evaluate Genome Level Transcription Papers

      Two Examples of "Alternative Splicing"

      Genes and Straw Men

      The Frequency of Alternative Splicing

      The Deflated Ego Problem

    2. Thanks Larry, I will check those out.