Testing Fluorescent Protein Sequence Entropy for Correlation with Protein Properties

Physics, specifically the Second Law of Thermodynamics, governs protein folding. The resolution of the Levinthal Paradox shows that proteins do not explore all possible configurations randomly; instead they collapse into intermediate states and/or assume their lowest energy native states. The primary sequence dictates the final native states, but there is no solid theoretical framework for identifying the most important amino acids necessary for folding. Currently, multiple sequence alignments of all members of a protein family are assumed to be structurally/functionally important. Validation of such assumptions requires cumbersome computational analysis. Here, we present our initial findings on the relationship between amino acid sequence entropy and observable properties of fluorescent proteins. We used the centralized database FPbase of fluorescent protein (FP) characteristics to facilitate discovery and inferences about FP properties. The FPbase sequence analyses were done as follows: 1) sort FP by length and remove outliers, 2) align FP sequences, and 3) perform Shannon entropy calculations to quantify sequence variability. We concluded that the amino acids essential for FP structure and function typically have low entropy values. This research represents a significant step towards devising a statistical approach to identify key amino acids in FPs and improve our understanding of protein folding.