Calculating the expected frequencies of potential secondary structure in nucleic acids as a function of stem length, loop size, base composition and nearest-neighbor frequencies.

A formula is derived to estimate the expected frequency of potential secondary structure of a given stem length and loop size in a random sequence of nucleotides of known length and composition. An example is provided using the SV40 sequence which shows significant non-randomness in stem length but whose significance is reduced by more than six orders of magnitude by taking nearest neighbor frequencies into account.