Fig. 2
From: Predicting protein-binding regions in RNA using nucleotide profiles and compositions
Structure of a feature vector. For a sequence of n nucleotides, mPWM and dPWM are represented by n and n−1 elements, respectively. Compositions represent the frequency of each mono-nucleotide (4 elements), di-nucleotide (16 elements) and tri-nucleotide (64 elements) in the RNA sequence. A protein sequence is represented by 63 elements (7 compositions, 21 transitions and 35 distributions)