Replies

I don't think it's as much of a quandary as has been made out. Edsheppa talked briefly about frame shifts - just with a frame shift, a single gene can potentially code for three different proteins. So even if it turns out that we "only" have 30,000 genes or so, if each gene pulls triple duty, that would come close to explaining the apparent "gap" between the number of genes and the complexity of the organism.

The question to be settled is whether any finite list can predict outcomes. Is DNA a blueprint, and can entirely novel organisms be designed.

This is an empirical question, but an unsettled one.

My skepticism is based on experience with software design. With computers you have an environment in which every possible instruction has a completely defined effect. Yet it is difficult to design software that is reliable. Even worse, it isn't possible with the current state of technology to automatically translate a program from one delelopment language to another, or automatically port a software system from one Operating System or CPU to another. These tasks all involve finite lists of operations, but are too complex to automate.

I'm not convinced that it is possible to predict all the interactions that occur when DNA is "decoded".

It has been more than a century since Darwin -- who had no concept of genetics or DNA -- and the best gusee as to how life works is still variation+overproduction+selection. This is a pretty sloppy way for a computer program to operate, hence I maintain there is a fundamental difference between DNA and current computer programs.

Edsheppa talked briefly about frame shifts - just with a frame shift, a single gene can potentially code for three different proteins. So even if it turns out that we "only" have 30,000 genes or so, if each gene pulls triple duty, that would come close to explaining the apparent "gap" between the number of genes and the complexity of the organism.

The genotype-phenotype map that js1138 is talking about is something completely different. A frameshift mutation changes the genotype but does not change the relationship it has to phenotype. Phenotype is determined by a number of post-transcription events such as RNA-editing, alternative exon combinations, post-translational alterations, etc.