From: Matthew P Wiener To: All Jan-31-94 01:47PM Subject: Re: Evolution is WRONG!!! Organization: The Wistar Institute of Anatomy and Biology From: weemba@sagi.wistar.upenn.edu (Matthew P Wiener) Message-ID: <2ijua0$76p@netnews.upenn.edu> Reply-To: weemba@sagi.wistar.upenn.edu (Matthew P Wiener) Newsgroups: talk.origins ------------------------------------------------------------- Evolution makes numerous predictions. For example, plate tectonics and radioactive dating make it very clear that the Hawaii chain of islands gets older as one heads northwest from the big island. So evolution makes a great big prediction here: the various species will be distributed over the islands in such a way that the ones on Honolulu will be more closely related to the mainland species while the ones on Hawaii will be more closely related to each other than to the ones on Honolulu. One can make this more precise over all the islands. Care to predict what investigating a hundred species of fruit flies has come up with? If you want numerical quantification, there are cladograms based on chromosome inversions, and dendrograms based on DNA sequencing. ------------------------------------------------------- The first and most famous of all predictions of evolution is usually not thought of as such, but as a prediction it's solid gold, and one of the most astonishing in all of the history of science. Lord Kelvin, the premier physicist of his day, objected to the time frame Darwin was describing in his first edition, based on Kelvin's best comprehension of the earth cooling from a molten state. (Contrary to popular opinion, Kelvin did believe in evolution. He just believed it was directed by the hand of God, and so went much faster than Darwin allowed for.) Since Darwin wasn't too concerned with the physics and geology, he silently dropped the time estimates. But the hard-core evolution-rules attitude, please leave God out of the science, gave a rather basic prediction: Kelvin was wrong, and something was missing in his equations. It turned out to be radioactivity. I can't overstate this. Physics then (and today) was the pinnacle of the sciences. An argument from physics carried far more weight than any argument from biology. Yet it turned out that the _physics_ was flat-out embarrassingly wrong. -------------------------------------------------------- Naked mole rats are another wonderful prediction. One of the most difficult problems with evolution, from Darwin on, is how to account for the social insects. How does one evolve sterility? Hamilton offered an explanation based on genetics in the sixties. It was based on the peculiar genetics of ants, bees and wasps: the males are an X, and the females an XX. Given this, a female is closer related to her sisters than her own daughters. But this explanation is unsatisfactory for termites, with normal genetics. In the mid seventies, Alexander offered an alternative ecological explanation for termite eusociality, based on their difficult choice of habitat. Because he took for granted that evolution is what made for termites, he could extrapolate what would happen to mammals in an analogous situation. He described in his lectures at the time what such an hypothetical mammal would have to be like: a hairless subterranean rodent living in east Africa, since only there would conditions be right for his version of termite evolution to carry through. And what do you know, after some time somebody recognized that animal he was describing, and so Alexander wrote a carefully worded letter to the world's leading (and only) authority on naked mole rats. All at once, the weird breeding habits of NMRs made sense, and people realized that NMRs were also eusocial. --------------------------------------------------------------- Evolution predicts thousands of transitional fossil forms would be found. Darwin staked his theory on this. They've been found in amazing quantity. See the [list]. ---------------------------------------------------------------- Of more recent vintage, the list is endless. One that caught my eye a few weeks ago, and which I posted here, was the question of asymmetric mutation rates in DNA strands. As is well-known, some programmers at some point realized that if the mechanisms of evolution often produce optimal or near optimal solutions to complicated problems, then this should be applicable as an optimization method. This basic prediction of evolution is now the field of "genetic algorithms". In fine-tuning the method, one asks, just how do mutations work? Standard implementations assumed mutation was symmetric. The PNAS paper I recently desribed investigated asymmetric mutation rates for the leading and lagging DNA strands during replication. They discovered that a superior optimization algorithm existed for a bin packing problem if one used different rates. The implicit prediction--that nature beat us to it--was verified by an indepedent research group. From the point of view of this independent research group, who found asymmetric mutation rates in actual DNA of actual cells, the prediction went the other way: viz, GA algorithms would be superior if they used asymmetric mutation rates. This work is eminently "numerical". ---------------------------------------------------------- A classic problem has been the nature and evolution of sex. All sorts of predictions abound here. For example, life is difficult for parasites if the descendents' immune system is as different as possible from either of its parents. Hence, the incest taboo. Evolution predicts that an actual biological, not social, mechanism for enforcing this would exist in all parasite prone species. It turns out that the smell of MHC-II (the key immunological recognition molecule) is a heavy factor in mate selection in mice. Take an unrelated mouse, and paint it with odor of MHC-II of a close relative, and the opposite sex loses interest. (For what it's worth, people too can actually discriminate MHC-II smells.) Of course, some parasites are known to mutate at high rates. Why in the world would they do that? The origin of sex is a popular topic. It seems, at first blush, to be half as efficient. OK, so you buy the recombinatino argument--changing genes around deliberately is good for the species in the long run. Why two sexes? Why not six? Well, multiple sexes are known in some fungi, but overall, they are rare. The reason seems to be the following: evolutionarily speaking, multiple sexes are unstable. There is a hierarchy of mating types, and a mutation that cheats vis-a-vis where it is in the hierarchy will survive. The long range effect (if you believe in evolution) is to cancel the hierarchy into just two sexes. So, evolution makes a definite prediction regarding some of these fungi. What's your predictions? -- -Matthew P Wiener (weemba@sagi.wistar.upenn.edu)