%0 Journal Article
%T Predicting dihedral angle probability distributions for protein coil residues from primary sequence using neural networks
%A Glennie Helles
%A Rasmus Fonseca
%J BMC Bioinformatics
%D 2009
%I BioMed Central
%R 10.1186/1471-2105-10-338
%X In this paper we develop an artificial neural network that uses an input-window of amino acids to predict a dihedral angle probability distribution for the middle residue in the input-window. The trained neural network shows a significant improvement (4-68%) in predicting the most probable bin (covering a 30¡ã ¡Á 30¡ã area of the dihedral angle space) for all amino acids in the data set compared to baseline statistics. An accuracy comparable to that of secondary structure prediction (¡Ö 80%) is achieved by observing the 20 bins with highest output values.Many different protein structure prediction methods exist and each uses different tools and auxiliary predictions to help determine the native structure. In this work the sequence is used to predict local context dependent dihedral angle propensities in coil-regions. This predicted distribution can potentially improve tertiary structure prediction methods that are based on sampling the backbone dihedral angles of individual amino acids. The predicted distribution may also help predict local structure fragments used in fragment assembly methods.The primary sequence of a protein is believed to define the three-dimensional (tertiary) structure of the protein and many attempts at predicting the tertiary structure from primary sequence has been made (see for instance [1] for an overview of the CASP VIII experiment).The main reasons that predicting protein structure from sequence alone is so difficult, is that the possible ways the amino acids can twist and turn with respect to each other are enormous. However, large parts of most proteins are arranged in secondary structures like helices and sheets, in which the dihedral angles of the amino acids lie within fairly limited areas as can be observed in Ramachandran plots [2-4]. Fortunately, predicting secondary structures can be done quite accurately [5-8], and since roughly 60% of amino acids in most proteins are arranged in these secondary structures [9], the number of possib
%U http://www.biomedcentral.com/1471-2105/10/338