The fidelity of the tag-antitag system


In the universal DNA chip method, target RNAs are mapped onto a set of DNA tags. Parallel hybridization of these tags with an indexed, complementary antitag array then provides an estimate of the relative RNA concentrations in the original solution. Although both error estimation and error reduction are important to process application, a physical model of hybridization fidelity for the TAT system has yet to be proposed. In this work, an equilibrium chemistry model of TAT hybridation is used to estimate the error probability per hybridized tag (ϵ). The temperature dependence of _ is then discussed in detail, and compared with the predictions of the stringency picture. In combination with a modified statistical zipper model of duplex formation, implemented by the Mjolnir software package, ϵ is applied to investigate the error behavior of small to moderate sized TAT sets. In the first simulation, the fidelities of (1) 105 random encodings, (2) a recently reported Hamming encoding, and (3) an ϵ-based, evolved encoding of a 32-strand, length- 16 TAT system are estimated, and discussed in detail. In the second simulation, the scaling behavior of the mean error rate of random TAT encodings is investigated. Results are used to discuss the ability of a random strategy to generate high fidelity TAT sets, as a function of set size and encoding length.

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)