Prediction and quality assessment of transposon insertion display data.

Vol. 36, No. 2 (2004) BioTechniques 223 Transposons are mobile sequences commonly found in prokaryotic and eukaryotic genomes. Their dispersal, repetitiveness, and the fact that their mobilization is a source of polymorphism make them choice candidates for use as molecular markers in mapping technologies. In recent years, variations of a technique inspired by amplified fragment length polymorphisms (AFLPs) (1) that take advantage of transposons have emerged as valuable tools for molecular analyses (2–7). These transposon-based mapping techniques, referred to as transposon insertion display (TID) after the first published report (2), have been applied to plants as well as to animals for population analysis (8,9), detection of transposition events (10), gene tagging (2), and the recovery of integration sites (3). The common basis of all TID techniques is an adaptor-mediated multiplex PCR amplification of genomic restriction fragments that contain a transposon marker sequence, along with the variable length of the DNA sequences flanking the insertion sites (Figure 1). However, to avoid the amplification of restriction fragments that do not contain transposon marker sequences (i.e., nonspecific), different adaptor designs have been adopted (Figure 1, A and B) (2–11). TID protocols also differ in the type of transposon chosen as a marker. However, transposon abundance, diversity, and distribution vary greatly between organisms. For example, transposon content can range from 3% in the yeast genome to over 60% in maize, and mammalian genomes primarily contain long and short interspersed nuclear elements, whereas the maize genome is mostly populated by long terminal repeat retrotransposons (12). This high variability can complicate the choice of an appropriate marker since clarity and resolution depend on the copy number of the transposon type used. Furthermore, the accuracy of TID is dependent on the design and PCR conditions of a transposon-specific primer. Transposons are characterized by structural features that may be problematic for PCR [e.g., terminal and subterminal repeats, secondary structures, A and T richness, or poly(A)/(T) tails]. Thus, the ability to predict the banding pattern generated by a specific primer in a specific genome would be useful for optimizing PCR conditions and for assessing the reliability and quality of the observed data. A large data set of genomic sequence is currently available, including the complete sequence for eukaryotic model organisms such as Arabidopsis, Caenorhabditis elegans, Drosophila, mosquito, rice, and human (http: //www.ncbi.nlm.nih.gov:80/PMGifs/ Prediction and quality assessment of transposon insertion display data