Theoretical evidence for the significance of certain types of DNA secondary structure in the genomes of phi X174, G4, fd, SV40 and the plasmid pBR322.

The significance of certain types of DNA secondary structure was assessed on a purely theoretical basis by analyzing their distribution over the entire genomes of bacteriophages phi X174, G4, fd, the eucaryotic virus SV40 and the E. coli plasmid pBR322. A computer program was designed to search the translated and non-translated (intercistronic) regions of these genomes for the potential of forming perfect hairpins (perfectly base paired stems: no G-T pairing; no interruptions) with small loops. The number of found occurrences was then compared to the statistically expected number. The results reveal a significant excess in each genome of perfect hairpins with stem length greater than 7 and end loops less than or equal to 20 over that expected given their nucleotide composition. For stem length of 5 to 7 the same observation holds true for the intercistronic regions, but for translated regions the number of such structures is below expectation. This suggests an evolutionary selection for these structures.