Friday, May 3, 2013

Revisiting DNA as a Computation Model

Just a brief note, perhaps as inspiration to get things organized.

The idea calls out to me again, and a little bit of digging around has surfaced hints from other researchers that:

1) DNA processes can be modeled as a Turing-Complete computing model. I need to find the actual papers that say this and read them in my free time.

2) Peptide-Antibody interactions are also Turing-Complete. Balan have done some work on this. I'll list these in a later update after I've read the papers.

I am interested to start getting information organized to investigate to what extent bio-chemical interactions can be treated as a Turing-Complete computation model, how far the idea can be taken for practical computing purposes, and what philosophical implications it has for the understanding life and its relationship to chemistry and particle physics.

At the lowest levels, I would like to know:

a. What are the input and output functions of such models of computation? In the case of DNA, if one treats RNA chains as the no-write instruction set, then one could imagine amino acids as the input and proteins as the output, with enzymes being something of that strange in-between which transforms the data for further input/output dynamics I current am nowhere close to understanding.

b. Under what conditions do these bio-chemical interactions work? Assuming water as the medium at first, under what temperatures and pressures can/should we apply our model? How does the chemistry change with changes in temperature and pressure? Do the rates change? Do the component (e.g., amino acids) concentrations change? How does one supply the energy necessary to effect the chemistry desired? Can these parameters somehow be analogous to CPU frequency changes in silicon-based computing chips?

c. What kind of timing mechanisms are employed by life's biochemistry? I have one paper (Howard & Gerdes) with some hints, and I'm hoping to find more. As I understand it, there is an entire orchestration of what are effectively Brownian processes that drive our biological functions. How does the system ensure the smooth operation of this orchestration and under what conditions? How do these timing mechanisms change under the conditions described in point b?

d. Way-out-there ideas - what happens in a medium other than water (e.g., liquid methane like on Titan)? Can other chemical processes involving related atoms (e.g., silicon for carbon, arsenic for phosphorus) hypothetically work? Do they require different energy levels for similar chemical interactions? Do they require a different medium? Can we also create clean-room environments to isolate water-based interactions for computational purposes? Real-life biology is full of other bio-chemicals that need to be accounted for, is my thought.

Finally, how should I get such information organized? Obviously, this blog is a poor media for it. I am thinking of Wikia, but I need to be comfortable with privacy settings and dealing with any intellectual property issues before proceeding. My primary goals are to figure out the landscape of the literature surrounding the topic. The first steps I have taken so far were to establish some baseline faith that bio-chemistry can in fact be made to conform to a Turing-Complete computation model - efforts to move to efficient computation models and implementations can follow from this baseline.

No comments:

Post a Comment