Figure 4
From: Metamotifs - a generative model for building families of nucleotide position weight matrices
The effect of metamotif position weight matrix prior on motif discovery with Nested MICA. A) Metamotifs used as the motif priors. Error bars represent 95% confidence intervals. B) Informative weight matrix prior improves NMICA's sensitivity to resolve motifs present in human intronic sequence in low frequency (0.2 frequency). The bars represent the sequence length at which a motif closely similar to the input motif was successfully recovered (p < 0.05, empirical p-value defined in [3]).