Theory refinement on Bayesian networks
Proceedings of the seventh conference (1991) on Uncertainty in artificial intelligence
Using a neural networking method to predict the protein phosphorylation sites with specific kinase
ISNN'06 Proceedings of the Third international conference on Advances in Neural Networks - Volume Part III
| Hi-index | 0.00 |
In the post-genome era, huge numbers of protein structuresaccumulate, but little is known about their function. It is timeconsuming and labour intensive to investigate them, e.g., enzymecatalytic properties, through in vivo or in vitro work. So insilico predictions could be a promising strategy to greatly shrinkthe list of potential targets. This work incorporated bothstructural and physico-chemical information into a Naive Bayesclassification system, and gained much better performance. Theten-fold cross validation results of this method could reach 88.6%of sensitivity and 93.7% of specificity. The improvement ofprediction accuracy is detailed in this paper. The PECB is alsoapplied to predict other important sites.