Figure 9

Benchmarks for different datasets. (a) Benchmarks for run time versus sequence length for sequences ranging from 100nt to 10000nt. A 10000nt sequence can be analyzed in less than a minute; the non-linear increase in run time as a function of sequence length is due to the number of secondary structures that are predicted (b) Accuracy versus sequence length. To ensure there was no loss of accuracy with longer sequences we also calculated the accuracy for the dataset to ensure there was no size effect.