Journal of Biomedical Informatics
SWARM: a scientific workflow for supporting bayesian approaches to improve metabolic models
CLADE '08 Proceedings of the 6th international workshop on Challenges of large applications in distributed environments
Max-margin Classification of Data with Absent Features
The Journal of Machine Learning Research
Comparison of Bayesian and regression models in missing enzyme identification
International Journal of Bioinformatics Research and Applications
A Bayesian Approach to High-Throughput Biological Model Generation
BICoB '09 Proceedings of the 1st International Conference on Bioinformatics and Computational Biology
RCG'06 Proceedings of the RECOMB 2006 international conference on Comparative Genomics
An evolutionary approach for searching metabolic pathways
Computers in Biology and Medicine
Hi-index | 3.84 |
Motivation: The metabolic models of both newly sequenced and well-studied organisms contain reactions for which the enzymes have not been identified yet. We present a computational approach for identifying genes encoding such missing metabolic enzymes in a partially reconstructed metabolic network. Results: The metabolic expression placement (MEP) method relies on the coexpression properties of the metabolic network and is complementary to the sequence homology and genome context methods that are currently being used to identify missing metabolic genes. The MEP algorithm predicts over 20% of all known Saccharomyces cerevisiae metabolic enzyme-encoding genes within the top 50 out of 5594 candidates for their enzymatic function, and 70% of metabolic genes whose expression level has been significantly perturbed across the conditions of the expression dataset used. Availability: Freely available (in Supplementary information). Supplementary information: Available at the following URL http://arep.med.harvard.edu/kharchenko/mep/supplements.html