Mean and variance of the Gibbs free energy of oligonucleotides in the nearest neighbor model under varying conditions.

Abstract

In order to assess the stability of DNA-DNA hybridizations-for example during PCR primer design or oligonucleotide selection for microarrays-one needs to predict the change in Gibbs free energy DeltaG during hybridization. The nearest neighbor model provides a good compromise between accuracy and computational simplicity for this task. To determine optimal combinations of reaction parameters (temperature, salt concentration, oligonucleotide length and GC-content), one would like to understand how DeltaG depends on all of these parameters simultaneously. We derive analytic results about the distribution of nearest neighbor DeltaG values for a Bernoulli random sequence model (specified by oligonucleotide length and average GC-content) under given experimental conditions. We find that the distribution of DeltaG values is approximately Gaussian and provide exact formulas for expectation and variance.