Instruction, selection and genetic program :Ramellini : Wed, 01 Dec 1999 16:51:59 GMT

Instruction, selection and genetic program

According to Kupiec and Sonigo (1999, see HBG essays), the correspondence between DNA and polypeptide sequences can be explained or, better, interpreted either by instructive or selective models.

The observative data show that more than one aminoacid sequence correspond to one DNA sequence (seldom, usually or always? This is not clear to me). This observation is interpreted in a different way by the two alternative models.

The instructive model speaks about errors in protid synthesis: there is one correct protid and many uncorrect or illegitimate products; the criterion to distinguish among correct and uncorrect products is the linear correspondence between certain DNA triplets and certain aminoacids, according to the rules of the genetic code. The selective model says that all the products of the synthesis are biologically significant, and that a sort of broad selection among them takes place; so, according to this model, there are no univocal or specific instructions.

What about the genetic program, seen as a system of instructions?

This concept is pleasant to the instructive model while it seems rejected by the second one. Nevertheless, I think that the analogy of instruction could be accepted by both the models, though in a different way: the instructive model conceives instructions as iron obligations, the selective one as kindly suggestions. If it so, the difference between the two models wouldn't be the presence/absence of instructions, but the compulsoriness of instructions. The instructive model would be better called a univocal, specific instructive model, while the selective one would be a equivocal, unspecific instructive model.
RE:Instruction, selection and genetic program : satamas : Thu, 02 Dec 1999 01:39:01 GMT

You message is so much saturated with important statements and questions, that I was going to reply with a big letter, but because of time limitations, let me refer to several relevant web sites:

http://post.queensu.ca/~forsdyke/bioinfor.htm

http://home.wxs.nl/~gkorthof/kortho37.htm

I think Pierre Sonigo and Jean-Jacques Kupiec will answer in details and with better explanations.

Generally speaking, I believe that Kimura's neutralism is very relevant here. By the way, can anyone explain to me why Kimura is considered an anti-Darwinist?

RE:Instruction, selection and genetic program : sonigo : Sun, 05 Dec 1999 20:07:16 GMT

Difficult to answer in a few lines to these difficult questions. May be a book :)

(We are writing a book with JJ Kupiec)

A few elements to Pietro : switching from instruction to "suggestion" has huge consequences.

When the instruction is compulsory (true instructive model), the future is written in the past. The instruction is causal. This is a problem (amongst many) for the evolutionary origin of the instruction.

When the instruction is only a suggestion (selective model), and this suggestion originates from random variations, nothing pre exists but a "noise" full of potential. Before selection, all suggestions ("valid" or not) have the same status : undefined variations. Their status can be known retrospectively. Knowing the possible "suggestions" exhaustively in such a situation does not really help : the system of selection is not written in the set of random suggestions. The preexisting set of suggestions is not causal. The selection, here and now, is causal.

"Dans un modele instructif, l'information precede l'existence, dans un modele selectif, elle lui succede" (Philippe Lherminier)

Concerning Kimura, a few elements for Sergei (too schematic, I admit):

Original darwinian theory : evolution is due to natural selection acting on phenotypes

The problem was to introduce the genotype in the theory.

Hypothesis : selection causes variation of allele frequencies

Observation and mathematical models : variation of gene frequencies can be independent of selection

Surprisingly, instead of refuting the hypothesis, the initial theory was changed, leading to neutralism.

Neutral theory : evolution can be independant of selection. Variation of gene frequency became the new definition (and cause) of evolution. In this context, random variations of gene frequencies are also considered as evolution, which means evolution is driven by chance only. Without selection, this is indeed a non-darwinian theory.

The origin of such an epistemological mess is trying to reconcile an instructive model of biology (genetics) with a selective one (evolution). The neutral theory was a victory of genetics, leading to a dramatic destruction of darwinism. On the contrary, the victory of evolution implies a revision of genetics. This is where we try to progress with our essay : from genotype to phenotype, instruction or selection ?

What do you think ?