September 25, 2016 at 03:17AM

Today I Learned: 1) For a simple model of genome evolution, there is a critical mutation rate above which a species cannot maintain selective traits. The intuition is that it is much easier to mutate away from a specific genome than it is to mutate back to it, so if the overall mutation rate is high enough, it can overwhelm any selective advantage of the most-fit genome -- even if the most-fit genome has more offspring, too many will be less-fit mutants for that fitness advantage to matter. The transition is, at least for very sharply peaked fitness peaks, an abrupt phase-change. Below that mutation rate, the species will maintain a stable equilibrium of variants spread around the peak, centered on the peak. Above the threshold, adaptation more or less completely breaks down, and the species will explode in genetic diversity. The critical mutation rate threshold is, moreover, proportional to 1/N, where N is the genome length -- the larger the genome, the less mutation it can tolerate before becoming essentially unstable over multiple generations. I'd heard this idea before (that small genomes are in some sense more tolerant to mutations). Today I saw the math behind it. I won't go into it here, but suffice it to say that it's pretty simple, and pretty simplified. I also learned today that the critical mutation rate turns out to be close to 1 mutation per genome per generation, and it turns out that many species (dare I say most?) do fall under the regime of <1 mutation/genome/generation. Also, today I learned of an interesting side effect of the above stuff -- the mutation rate threshold is bigger for more broadly-distributed peaks in fitness space. That means that if there are two peaks in fitness space, with one very sharp and tall and the other less fit but broader, then a species with moderate mutation rate may actually gravitate to the broad, shorter peak instead of optimizing for global maximum fitness. I find that very satisfying, because it's a very formal way of thinking about "evolutionary robustness", which comes up fairly frequently in conversation about biology but is a little hard to justify intuitively. 2) Related to the above, think about how you would guess the following organisms rank in terms of mutation rate per genome *per generation*: human RNA viruses; bacteriophages; E. coli; yeast; fruit fly; mouse; human. Which do you think has the highest per-generation mutation rate? If you guessed "mouse", you're close -- estimates put mouse per-generation mutation rates at about 1/2 a mutation per genome per generation. But the highest rates in the above list belong to lytic RNA viruses. The table I'm reading from has per-generation mutation rates between 0.84 and 6.5, which is shocking above the critical mutation rate mentioned before. Which do you think has the lowest per-generation mutation rate? It's bacteriophages. Turns out that with the glaring exception of RNA viruses, the per-genome mutation rate is dominated by genomic size. Bacteriophages have relatively high per-base mutation rates (1,000-10,000 times higher than humans) but their tiny size means that their per-*genome* mutation rate is quite low (around 0.004/generation). 3) The shape of the clitoris (which, in case you don't know, is a much larger organ than it appears from the outside, and extends around much of the vagina) was discovered fairly recently. See this *1998* paper descibing the first modern dissection of the clitoris: http://ift.tt/2d0Bvzb. Here's the really obnoxious bit -- the shape of the clitoris *was* well known before the 20th century. Early versions of Grey's Anatomy included pretty accurate anatomical structures of the clitoris. Later versions, and virtually all other medical literature and textbooks, simply omitted them until fairly recently.