Abstract
Characteristic peptides of the protein segments having common secondary folds are obtained for the I-sites library using maximal position specific probability scores. The secondary structures of these peptides are predicted deploying two best-known computational methods. These are validated with significant accuracy against the corresponding motifs. The characteristic peptides also match with those computed using a Bayesian modeling approach with Markov Chain Monte Carlo Simulation. Percentage representation of the characteristic peptides in the protein structural and functional families shows some interesting results with potential applications in protein structural genomics.
Keywords: I-Sites library, peptides, position specific probability, probabilistic characterization, protein secondary structure, sequential motifs