Derivation of nearest-neighbor properties from data on nucleic acid oligomers. II. Thermodynamic parameters of DNA.RNA hybrids and DNA duplexes.

Using nearest-neighbor models consisting of independent short sequence combinations of nearest neighbors (ISS models), values of thermodynamic parameters for sets of independent sequences are derived from published oligomer data for DNA.RNA hybrids [N. Sugimoto, S. Nakano, M. Katoh, A. Matsumura, H. Nakamuta, T. Ohmichi, M. Yoneyama, and M. Sasaki (1995) Biochemistry, Vol. 34, pp. 11211-11216] and dsDNA duplexes [J. SantaLucia, Jr., H. T. Allawi, and P. A. Seneviratne (1996) Biochemistry, Vol. 35, pp. 3555-3562]. The results are compared with those from models that assign values of thermodynamic parameters to individual nearest neighbors (INN models). Differences in the use of ISS and INN models are also illustrated in an appendix, which shows examples of analyses for values of a fictitious nearest-neighbor property. INN models that include an initiation parameter contain an implicit assumption that combinations of end neighbors have the same value of a property. It is found that combinations of end neighbors (e.g., base pairs neighboring solvent) in oligomers can have significant and different apparent values of thermodynamic properties, so that the assumption inherent in INN models is not always correct. Even though ISS models do not allow the assignment of values to individual nearest neighbors, except for the like neighbors [such as d(AA)/r(UU), etc., for hybrids and d(AA)/d(TT) and d(GG)/d(CC) for DNA duplexes], they do provide physically meaningful values for the like neighbors, for sequence combinations, and for specified combinations of end neighbors.