Does anyone know how far the decoding of morphological data in the DNA has reached ? (e.g. : can one find how many legs a given creature has, only by looking at the DNA code ?)

Not as far as I know. Building legs and such is a very complex process, involving hundreds or even thousands of genes.

To your question no, AFAIK the most we can do with a sequence is compare it to the many others and find out the likely function of it if it is an enzyme or structural protein, however if the data isn't present then that is pretty much the end point you can get from the sequence, you can even tell the 3D structure.

The 3D structure of the protein ?

The 3D structure of the protein ?

 

Genes encode the linear amino acid sequence. As the protein is translated, it starts to fold up into a particular shape (or a set of shapes), due to the fact that some amino acids are more hydrophobic than others. Think of that as the "oil drop" model, the amino acid chain balling up with all of the hydrophobic amino acid residues on the inside and all of the hydrophilic residues on the outside (exposed to water).

 

Actual protein structure is more complicated, often characterized by particular structures like alpha-helices, beta-pleated sheets, immunoglobulin folds, barrels, etc. Some structures are stabilized by hydrogen bonds between side chain and/or backbone atoms, while others are stabilized by C-C disulfide links. And then there is the chaperonin class of proteins, which fold or refold proteins.

Is it true that Genes represent 1% of DNA ? ....but anyhow, this implies that DNA could not contain any macroscopical data encoded ?

I don't know about 1%, but there is an awful lot of DNA that's not used to code proteins. As for "macroscopical data", I'm not sure what you mean.

By "Macroscopic data" was meant : DNA data coding for phenotype (e.g. number of legs, relative size of organs) when mature.

Oh, DNA definitely codes for phenotype, such as number of legs, size of organs, etc. It's just a very, very complex process involving hundreds of genes and regulatory elements.

 

Essentially, your arm is the same as a bat's arm - what differs is when the genes are turned on and off and at what concentrations, which is controlled by either developmental genes or non-gene regulatory sections of the DNA.

With regards to the coding sequences: according current ORF (open reading frames) estimations roughly 1-2% of human DNA encode proteins. A sizeable but yet not closely mapped amount will also code for non-coding RNAs (especially the realm of small RNAs is under investigation). And of course a lot of non-coding DNA will be involved in regulatory functions, as Mokele hinted at.

And how do researcher find out which genes act or regulate on phenotype; is there already database ? : is there another way controlled mutations ? But this is ethically not possible for human, so this database could not exist.

In most cases data is obtained by model organisms (e.g. certain bacteria, fruit flies, mice etc.) from which mutants are created. The results of altering or, more commonly, inactivation of certain genes is monitored. Obviously this is not done with humans. Generally the animal model is applied to humans and in some cases genetic diseases provide additional clues.

It's also worth noting that we've got a pretty good understanding of what mutations to various genes do to humans. Yes, we can't deliberately induce mutations, but there's a hell of a natural sample size (6,500,000,000 and growing) with lots of natural mutations, and we pay a *lot* more attention to medical defects in humans for obvious reasons (plus they can pay the doctor). As a result, we're got a pretty good accumulation of knowledge from just noticing congenital abnormalities and figuring out the reason.