bill@utastro.UUCP (William H. Jefferys) (05/02/84)
Responding to an article by Dave Norris which first appeared in net.religion: >The following is taken from Boa & Moody's "I'm glad you asked." I offer it >in the hope that some informed soul can shed some light on its veracity... > ... let us assume an ideal environment with a primordial soup >full of amino acids and the proper catalysts, with just the right temperature >and moisture. Some estimate that under these favorable conditions the >chances of getting dipeptides (two amino acids bonded) would be about 1 in 100. >But the chances of tripeptide formation would be about 1 in 10,000. To get a >polypeptide of only 10 amino acids, the probability would be 1 chance in >100,000,000,000,000,000,000 (100 quintillion). Yet the proteins in the >simplest living things have chains of at least 400 amino acids on the average. Who made these estimates? Without a reference one cannot check to see what the assumptions were. Anyway, there is an excellent chance that the precursors of life were not proteins at all but rather self-replicating strands of RNA (R. Lewin, 1982. *Science* 218:872-874). Laboratory experiments have, under the right conditions, even generated self-replicating RNA sequences *de novo* (that is, without an initial template to model the sequences on) (M. Eigen et. al., 1981, *Scientific American*, April 1981, 88-118). These experiments show that the *de novo* synthesization of RNA is not at all random in that certain sequences are strongly favored over others. > To make matters worse, all proteins are built of amino acids that are >exclusively "left-handed" in their molecular orientation. Left-handed and >right-handed amino acids are mirror images of each other, and the chances of >formation are about the same. Although both kinds can link with each other, >the first living systems must have been built with left-handed components >only. Some scientists have evoked natural selection here, but this only >applies to systems that can already reproduce themselves. Without an >intelligent ordering agent, we have only chance to explain this amazing >phenomenon. For a chain of 400 left-handed amino acids, the odds would be >roughly equivalent to tossing an ordinary coin and coming up with tails 400 >times in a row. The chances for that would be approximately 1 in 10^120 >( a 1 followed by 120 zeroes). All this for *one* protein molecule, >and hundreds of similar molecules would be needed in the first living >system. Not so. Laboratory experiments dating back even to 1940 show that (1) LL and DD polymers are much more likely than LD polymers. (L = left-handed and D = right-handed). The stereochemistry of polypeptides seems to be involved. (2) LD polymers, once formed, are much less stable than polymers consisting of all L or all D. (3) Even when they are formed, LD polymers are much shorter than all L or all D. (*Molecular Evolution and the Origin of Life*, Revised Edition, by S. Fox and K. Dose, Dekker 1977.) > None of this accounts for the fact that the 20 kinds of amino acids >operate like letters in an alphabet, and they must link in a meaningful >sequence to form a usable protein. A random sequence of amino acids >would be utterly useless. This fails to consider what *proportion* of random sequences *would* be useful. The evidence is that the proportion is rather high. First, typical proteins having the same function in different organisms (e.g., hemoglobin) have only a few sites that are *always* occupied by the same amino acid. In the case of hemoglobin, for example, only 7 of 140 sites are invariably occupied by the same amino acid (H. M. Perutz and H. Lehmann, 1968: *Nature* Vol. 219, p. 902). Moreover, examples are known where the *same* gene codes for *two different* proteins, by reading the gene shifted by one "letter". This is evidence that functional proteins generated from random genes are quite common. Anticipating what's below, how many speeches out of Shakespeare could be parsed by shifting over one letter to produce an entirely different meaningful speech? > DNA is far more complex than any of this, and it too is built out of a >highly organized alphabet. The letters are molecules called nucleotides. >A cell contains a chain of about three billion pairs of these nucleotides >(each gene has about 1,200 nucleotide pairs). The order of these >nucleotides or bases is crucial because every triplet of bases along this >immense chain is a word. Each word stands for one of the 20 kinds of >amino acids. Using these words the DNA can literally create any kind of >protein the cell needs. Some genes are much shorter. Certain organisms (Viroids) are known which contain only about 360 neucleotides, or 180 base pairs.i (T. O. Diener, *Scientific American* January 1981, pp. 66-73). The earliest self-replicating organisms were probably very small. > The amount of time required to synthesize even one gene (a paragraph of >these words) has been calculated by some scientists using absurdly >generous assumptions. Using a variation on a well-known illustration, >suppose a bird came once every billion years and removed only one atom from >a stone the size of the solar system. The amount of time required for the >stone to be worn to nothing would be negligible compared to the time needed >to create a useful gene by chance, even accounting for chemical affinities >and an ideal environment. Shaw's monkeys would long since have pounded >out the words of Shakespeare!" What scientists have made this calculation? Without a reference one cannot check to see what the assumptions were. >Earlier in the chapter, it was noted that the time required for 1 million >monkeys typing at 100 words/minute, 24 hours a day on typewriters with >40 keys to type the first four words of a Shakepearian play would be about >100 billion years. Here's a game you can play to see how this analogy fails: Take out the old Scrabble game, and turn the letters upside down (eliminate the "wild cards"). Mix them up. Start turning them rightside up one after another, and whenever several successive ones make a real English word, or sentence fragment, set them aside. You will be surprised how quickly you will get positive results. Now the odds against your having gotten *that particular word or sentence fragment* are extremely high. If you did the experiment again, you would never get the *same* string in a month of Sundays. Yet there are so many possibilities that in actual fact one is very likely to get *some* positive result. The point is, the argument assumes a false premise, that one has to make *one of a very small number of strings* (of RNA, DNA or amino acids). What is not mentioned is that a *high proportion* of strings must generate biologically active polymers, as was pointed out above. If only one in 10^9 strings of 100 elemental units is self-replicating, for example, then the probability of forming *one* of that set is extremely high. Under assumed primitive Earth conditions, with very restrictive assumptions about the polymerization rate, times of the order of a few decades have been estimated (S. Freske, *Creation/Evolution* Issue IV, pp. 8-16, Spring 1981). -- Bill Jefferys 8-% Astronomy Dept, University of Texas, Austin TX 78712 (USnail) {ihnp4,kpno,ctvax}!ut-sally!utastro!bill (uucp) utastro!bill@ut-ngp (ARPANET)