Biocomputing

[Corillian]

I didn't know where to put this so posted it here because of course, it is a computer.

How exactly do biocomputers work, i read an article about one in the UK but it didn't really explain much about how they work.

[5614]

I read about one recently too. The article didn't say too much, but from the sounds of it strands of DNA were used instead of silicon. Whilst the biocomputer worked, it was slow and not very useful in a practical sense, although they do show potential.

 

And that's about all I know. Google doesn't come up with much either. Hopefully someone else can expand on my answer.

[Dak]

well, the biological computation that is done by cells is usually done by phisical-feedback.

 

eg:

 

if you can transform protien P1 --> protien P2 using enzyme E, and you want x amount of P2, and the level of P2 that currently exists is z:

 

on a computer you could (using pretend-code) say something like:

 

get E to do (P1-->P2) x-z times #(x-z being the difference between the P2 you have, and the P2 you want)

 

or something like:

 

while z < x: #(untill the desired level of P2 is reached)

tell E to (P1-->P2) #(make P2)

 

either of which would result in transforming P1 into P2 untill the actual level of P2 (z) is the desired level (x).

 

as to how that works on an electrons-flowing-through-silicon-stuff level, that's beyond me.

 

in a cell, what you'd do, is youd desighn E so that P2 suppresses it by binding to it and changing it's shape, thus deactivating or slowing it down. if you balance the amount of E and the effect that P2 has on E properly, then the result should be that, as z approaches x, E will be increasingly inhibited, and the rate of E(P1 --> P2) will decrease.

 

when z = x (when you have enough P2), then E should be completely deactivated, so no more P2 is produced.

 

the effect is the same as the pseudocode examples: P1 is transformed into P2 untill you have the desired level of P2, at which point P2 production ceases.

 

similarly, if gene G makes enzyme E, you could desighn G so that E binds to and suppresses it. do it right (ie, get the amount by which E suppresses G right), and G will continue making E untill the correct amount of E is produced, at which point thered be so much E that G is totally silenced, and doesn't make any more E untill the level of E drops to a level that

 

1/ requires more E to be made to maintain desired levels, and

2/ is low enough so that G is no longer being totally silenced (see how it works?)

 

in cases such as the regulation of the cell-cycle, mitosis etc, this is actually quite complicated, with check-points that check certain conditions are met, and dont let the rest of the cell proceed untill they are (thus coordinating the whole thing, and making sure that, say, the chromosomes dont migrate to opposite poles of the cell untill they've been sorted into two groups, and that the cell doesn't divide into two untill the chromosomes are out of the way of the point at which the cell pinches into two -- which, coincidentally, is, through a chain of chemical interactions, guaranteed to be at right-angles to the poles to which the two chromosome groups migrate, thus ensuring that the pinching-into-two actually segregates the two groups of chromosomes).

 

the point is that cells, using just interaction between molecules, can process information to quite a high level of sophistication, and react to environmental 'input'.

 

if we can desighn a computer that works on molecules and that is programable (ie, useful), the idea is it'll be:

 

a/ really tiny -- smaller than current computers (good for data storage), and

b/ much faster in certain respects.

 

as for how contemporary man-made biocomputers work: i have no idea

 

tho, i think lamp-posts could be seen as a bio-computer (or at least a chemical-computer: not sure what the difference is). the pretend-code would be:

 

if light-level < x:

make light

 

else:

dont

 

the way that this works is that there are chemicals on top of the lamp-post that react to environmental light; when environmental light goes above x, the molecules are stimulated into inhibiting the lamp-post. when the environmental-light drops below x, the molecules become de-stimulated, and stop inhibiting the lamp-post, which procedes to make light. or something.

 

very very simple computation, but computation (and reacting to external input) non-the-less. the question is: how can we make them more complex, and prefferably programable.

 

hope that all made sence.

[ecoli]

http://technology-i.org/index.php?option=com_content&task=view&id=70&Itemid=33

 

I attending a lecture by this guy... Very interesting stuff.

[Corillian]

So its basically like the mini switches (can't remember what they are called) in a computer, but with cells and chemicals, i can see how that works. Isn't there a chance that it could evolve and become dangerous?

[insane_alien]

it can only evolve if it can reproduce itself. you are a biocomputer in some respects the only way YOU can evolve is to reproduce.

 

although there is the potential for insanity which is a whole other kettle of fish and is very likely to be dangerous.

[weknowthewor]

This explains the workings of the DNA-wave biocomputer in terms of a quantum mechanical theory called quantum holography..

 

http://www.rialian.com/rnboyd/dna-wave.doc