The use of the digital signal processing procedure of decimation is introduced as a tool to detect patterns of information entropy distribution and is applied to information entropy in influenza A segment 7. Decimation was able to reveal patterns of entropy accumulation in archival and emerging segment 7 sequences that were not apparent in the complete, undecimated data. The low entropy accumulation along the first 25% of segment 7, revealed by the three frames of decimation, may be a sign of regulation at both protein and RNA levels to conserve important viral functions. Low segment 7 entropy values from the 2009 H1N1 swine flu pandemic suggests either that: (1) the viruses causing the current outbreak have convergently evolved to their low entropy state or (2) more likely, not enough time has yet passed for the entropy to accumulate. Because of its dependence upon the periodicity of the codon, the decimative procedure should be generalizable to any biological system.
Information entropy (H) is a measure of uncertainty at each position within in a sequence of nucleotides.H was used to characterize a set of influenza A segment 7 nucleotide sequences. Nucleotide locations of high entropy were identified near the 5’ start of all of the sequences and the sequences were assigned to subsets according to synonymous nucleotide variants at those positions: either uracil at position six (U6), cytosine at position six (C6), adenine (A12) at position 12, guanine at position 12 (G12), adenine at position 15 (A15) or cytosine (C15) at position 15. H values were found to be correlated/corresponding (Kendall tau) along the lengths of the nucleotide segments of the subset pairs at each position. However, the H values of each subset of sequences were statistically distinguishable from those of the other member of the pair (Kolmogorov-Smirnov test). The joint probability of uncorrelated distributions of U6 and C6 sequences to viral subtypes and to viral host species was 34 times greater than for the A12:G12 subset pair and 214 times greater than for the A15:C15 pair. This result indicates that the high entropy position six of segment 7 is either a reporter or a sentinel location. The fact that not one of the H5N1 sequences in the dataset was a member of the C6 subset, but all 125 H5N1 sequences are members of the U6 subset suggests a non-random sentinel function.
The sentence sixth line from the end of paragraph two on page 355, “The second synonymous mutation was another G=>A transition at position 600 that converted the CAG codon to CAA, without change of encoded amino acid.”