Alexandre de Brevern - Thèse de Bioinformatique Moléculaire

abstract

Using an unsupervised cluster analyser, we have identified a local structural alphabet composed of 16 folding patterns of five consecutive C ("protein blocks"). The dependence that exists between successive blocks is explicitly taken into account. A Bayesian approach based on the relation protein block-amino acid propensity is used for prediction and leads to a success rate close to 35%. Sharing sequence windows associated with certain blocks into "sequence families" improves the prediction accuracy by 6%. This prediction accuracy exceeds 75% when keeping the first four predicted protein blocks at each site of the protein.

de Brevern AG, et Hazout S (2000),"Hybrid Protein Model (HPM): a method to compact protein 3D-structures information and physicochemical properties", IEEE - Computer Society : Proceedings of the 7th Symposium on String Processing and Information Retrieval, pp.49-54 [36].