The fidelity of the tag-antitag system II: reconciliation with the stringency picture

An equilibrium model was reported for predicting the hybridization error of a tag-antitag (TAT) system in response to a dilute input [J. Rose, et al., in DNA Computing, (Springer, 2001), 138]. One notable aspect of the model was its disagreement with the error behavior expected via a consideration of the melting temperatures (T'/sub m/s) of planned and error TAT duplexes, in isolation (the stringency picture). In this work, an equilibrium approach is applied to model TAT system error in response to a general, single-tag input. This model is implemented, using two-state and statistical zipper models of duplex formation, via Mathematica/spl trade/ and NucleicPark, respectively. Under appropriate nondilute conditions (high unplanned sequence-similarity; equimolar/excess input), an error minimum is predicted between the T'/sub m/s of planned and unplanned TAT duplexes, as expected via stringency considerations, providing a reconciliation between the two models of hybridization fidelity. On the other hand, predictions for a dilute input agree closely with the model in Rose, et al., supporting the view that a T/sub m/-based model is not fully adequate. A TAT set is then reported, evolved to support experimental validation, which exhibits excess input, signal-to-noise ratios that uniformly span the dynamic range of current fluorescence measurement equipment.