Skip to main content
Figure 5 | BMC Bioinformatics

Figure 5

From: HHMMiR: efficient de novo prediction of microRNAs using hierarchical hidden Markov models

Figure 5

Data flow for hairpin extraction from the genome. The genome is first folded using windows of 500 nts with 150 nts overlap between consecutive windows. Hairpins are then extracted from the folded windows using the parameters described in the text. Hairpins are pre-processed into a suitable format for training/testing using the states shown in Figure 3 (L: Loop; E: Extension; R: miRNA; P: pri-miRNA extension). For the purpose of testing, the folded sequence is pre-processed into 2 lines of input representing the 2 stems of the hairpin. An example is given in Figure 2b.

Back to article page