Predicting Protein Secondary Structure Using Stochastic Tree Grammars
Machine Learning - Special issue on learning with probabilistic representations
Handbook of graph grammars and computing by graph transformation: volume I. foundations
Handbook of graph grammars and computing by graph transformation: volume I. foundations
IEEE Intelligent Systems
Stochastic Graph Transformation Systems
Fundamenta Informaticae - SPECIAL ISSUE ON ICGT 2004
A glimpse of symbolic-statistical modeling by PRISM
Journal of Intelligent Information Systems
Stochastic Graph Transformation Systems
Fundamenta Informaticae - SPECIAL ISSUE ON ICGT 2004
| Hi-index | 0.00 |
The rapid progress in proteomics has generated an increased interest in the full characterization of glycoproteins. Tandem mass spectrometry is a useful technique. One common problem of current bioinformatics tools for automated interpretation of tandem mass spectra of glycoproteins is that they often give many candidates of oligosaccharide structures with very close scores. We propose an alternative approach in which stochastic context-free graph grammars are used to model oligosaccharide structures. Our stochastic model receives as input structures of known glycans in the library to train the probability parameters of the grammar. After training, the method uses the learned rules to predict the structure of glycan given a composition of unknown glycoprotein. Preliminary results show that integrating such modelling with the automated interpretation software program, GlycoMaster, can very accurately elucidate oligosaccharide structures with tandem mass spectra. This paper describes the stochastic graph grammars modelling glycoproteins.