Carbon — the first frontier of information processing

Abstract

Information is often encoded as an aperiodic chain of building blocks.Modern digital computers use bits as the building blocks, but in general the choice of building blocks depends on the nature of the information to be encoded. What are the optimal building blocks to encode structural information? This can be analysed by substituting the operations of addition and multiplication of conventional arithmetic with translation and rotation.It is argued that at the molecular level, the best component for encoding discretised structural information is carbon. Living organisms discovered this billions of years ago, and used carbon as the back-bone for constructing proteins that function according to their structure. Structural analysis of polypeptide chains shows that an efficient and versatile structural language of 20 building blocks is needed to implement all the tasks carried out by proteins. Properties of amino acids indicate that the present triplet genetic code was preceded by a more primitive one, coding for 10 amino acids using two nucleotide bases.Keywords. Amino acid; aminoacyl-tRNA synthetase; computation; genetic code; information; lattice models; protein structure; quantum search; tetrahedral geometry Structural informationIt is a characteristic of living organisms to acquire information, interpret it and pass it on, often using it and refining it along the way. This information can be in various forms or languages. It can be genetic information passed on from the parent to the offspring, sensory information conveyed by the sense organ to the brain, linguistic information communicated by one being to another, or numerical data entered in a computer for later use. It is advantageous to process the information efficiently, and not in any haphazard manner. In case of living organisms, Darwinian selection during evolution can be considered the driving force for such optimisation. In general, information processing is optimised following two guidelines: minimisation of physical resources (time as well as space), and minimisation of errors.A striking feature of all the forms of information listed above is that the messages are represented as aperiodic chains of discrete building blocks. Such a representation, called digitisation of the message, is commonplace due to its many advantages. Discretisation makes it possible to correct errors arising from local disturbances, and so it is desirable even when the underlying physical variables are continuous (e.g. voltages and currents in computers). It is also easier to handle several variables each spanning a small range than a single variable covering a large range. Any desired message can then be constructed by putting together as many as necessary of the smaller range variables, while the instruction set required to manipulate each variable is substantially simplified. This simplification means that only a limited number of processes have to be physically implemented leading to high speed computation. An important question, therefore, is to figure out the best way...

The Legacy of Alladi Ramakrishnan in the Mathematical Sciences

“…In this connection, it may be recalled that using the metaphor of an arch of stones, Cairns-Smith had proposed that the scaffold paving the way for "organic takeover" (the "arch") may well have been provided by clay minerals that were eventually disposed off. Indeed, this idea finds a sort of echo in the suggestion of Patel [121], viz., the choice of carbon with its tetrahedral geometry provide the simplest discretization of the fundamental operations of translation and rotation needed for processing structural information. (Rotations in 3-D are not commutative, a fact of crucial importance in representing structural information; in mathematical jargon this goes by the name of the SU(2) group of Pauli matrices/quaternions).…”

Towards Understanding the Origin of Genetic Languages

“…Finding the minimal language for proteins is a straightforward problem in classical geometry [Patel (2002)]. The following is a rapid-fire summary of the analysis.…”

Why genetic information processing could have a quantum basis

“…It is quite plausible that early forms of life existed with proteins that were made up of just 10 amino acids, belonging to one class or the other and coded by two nucleotide bases. The wobble rules and similar codons in the genetic code for amino acids with similar properties would then be relics of the merger of the two distinct classes during evolution (Patel 2001b). N = 10 is not an exact solution for Q = 2, which means that the quantum algorithm will not always find the desired object.…”

“…The building blocks of proteins-the amino acids-have to therefore know how to fold one-dimensional chains into three-dimensional structures. The number of building blocks required to encode one-dimensional chains and three-dimensional structures is certainly different (Patel 2001b). It is this drastic change in physical realisation of the information that has driven the living organisms to develop two distinct languages, and the complex machinery that translates one into the other.…”