On 15 Apr 1995, G. Dellaire wrote:
> In article <Pine.3.07.9504140953.B28952-e100000 at dev.gdb.org>,
>rrobbins at GDB.ORG (Robert Robbins) wrote:
>> >Taken all together then, the expression of the human genome involves the
> >simultaneous expression and (potential) interaction of something probably
> >in excess of 10**18 parallel processes.
>> Just curious, but how do you account then for the fact of imprinting and
> silencing of those parallel sets of instructions (i.e. alleles) such that
> you only actually have one "program" running. This would bring to light
> the idea of gene dosage, and what analog in informatics would pertain to
> such a mechanism?
Note that I was referring to "processes", not to programs. One program
on disk can result in many simultaneous processes being launched on a
parallel machine. Indeed, I would consider each mRNA to be a process, and
each active enzyme also a process.
With regard to the selective silencing of some copies of some programs,
whether because of imprinting, or dosage compensation, or differentiation,
or some other cause, these could be interpreted simply as examples
of the behavior of the 10**13 virtual machines (i.e., cells) in which the
programs on "disk" are being executed.
> On this point, it is becoming widely accepted that the actual structure of
> genome and not just the linear sequence may "encode" sets of instructions
> for the "reading and accessing" of this genetic code. Best illustrated
> by large changes in genomic structure that affects the accessiblity of
> of various regions of the genome to be "read" during development at specific
> times and silenced afterwards.
>> Therefore, a second level of language is the overall code itself. Sort of
> like the letters make up words (commands for the program), and different
> gene are like sentences... thus context is important for understanding the
> code. The context is provided by the supporting paragraphs which could represent
> genetic domains with in the genome (context can be spatial, what tissue;
> or temporal, what time of development). And the overall story provides an
> impression and message in itself.
This is a good point and demonstrates the multiple levels of subtlety and
conflation that must be taken into acciunt when trying to do linguisitic
analysis of DNA sequences. The storage medium, DNA, has local chemical
and mechanical properties that are affected by the information encoded in
those local regions (melting points, etc.).