The Design of Life: Information Technology, "Knapsack" and "Synprops" Used to Engineer the Sequence of Amino Acid Residues in Proteins

TL;DR

This paper presents an in silico model linking protein properties to their structure, sequence, and function, using Hopfield networks to analyze amino acid sequences and their evolutionary and structural implications.

Contribution

It introduces a novel computational approach combining information technology, 'Knapsack', and 'Synprops' to engineer amino acid sequences and understand protein structure and evolution.

Findings

Ideal amino acid sequences act as attractors for protein structures

Hopfield networks can recall specific protein states

Substitution acceptance exhibits nonlinear, emergent behavior

Abstract

An In Silico model to relate the properties of proteins to the structure, sequence, function and evolutionary history of proteins is shown. The derived ideal sequences for amino acid residues in proteins can then be considered as attractors for structures of actual protein families and their functions. Hopfield networks1 can then act as GCMs (General Content Addressable Memories) when they are trained to recall unique pre-determined states when presented with information associated with that state. The tertiary structure of proteins is determined by structurally enforced interactions between residues and they help to elucidate common motifs exhibited in proteins. The acceptance (or rejection) of substitution on the protein backbone conforms to a nonlinear response and can lead to emergent behavior.