New Strategies for Evaluation and Analysis of SELEX Experiments

Figure 2

The selection and weighting of secondary structure information prior to tree creation is illustrated. The top row shows the sequence of interest. In the following rows, for each secondary structure predicted by the Vienna RNA tool RNAsubopt [23], a representation in dot-bracket notation is placed. Mutual matching pairs of brackets denote base pairs, whereas dots are standing for unpaired bases. Each structure 𝑥 is annotated with an energy value 𝐸(𝑥) and the calculated probability 𝑃(𝑥) used for weighting. The nonpaired subsequences, which will be inserted into the tree, are highlighted yellow. It is obvious that some subsequences are inserted multiple times and others are overlapping, which necessitates the mentioned weighting.