The MONSTER server was developed by Huzefa Rangwala and George Karypis at the Department of Computer Science & Engineering, University of Minnesota.
The prediction methods that it uses are based on supervised learning that employ support vector machines (SVMs) on PSIBlast computed sequence profiles. Evaluations on various datasets have shown that MONSTER's method outperform existing protein residue annotation approaches. Details of these methods can be found in the following papers:
- A Generalized Framework for Protein Sequence Annotation. Huzefa Rangwala, Christopher Kauffman, and George Karypis. In Proceedings of the Workshop on Machine Learning in Computational Biology held in conjunction with Neural Information Processing Systems, Whistler, BC, Canada. 2007
- YASSPP: Better Kernels and Coding Schemes Lead to Improvements in SVM-based Secondary Structure Prediction. George Karypis. PROTEINS: Structure, Function, and Bioinformatics, Aug 15;64(3):575-86, 2006.
- Improving Homology Models for Protein-Ligand Binding Sites. Christopher Kauffman, Huzefa Rangwala, and George Karypis. To appear in Proceedings of LSS Computation Systems Biology Conference, California. 2008
- TOPTMH: Topology Predictor for Transmembrane a-Helices Rezwan Ahmed, Huzefa Rangwala, and George Karypis. To appear in Proceedings of the European Conference in Machine Learning, Belgium. 2008
- PROSAT: PROtein reSidue Annotation Toolkit Huzefa Rangwala, Christopher Kauffman, and George Karypis. Under Review.
Primary funding for this research was provided by the National Science Foundation (ACI-0133464) and by the National Institute of Health (RLM008713A). Access to computational resources is provided by the Minnesota Supercomputing Institute and the Digital Technology Center.
Additional information about this and other related research projects can be found at the Karypis Lab website.