Michael Wagner, PhD, works on applications of machine learning techniques to bioinformatics problems such as protein structure prediction, disease classification and protein identification. His research lab currently is investigating machine-learning based scoring algorithms for peptide mass fingerprinting to better understand how to optimally mine mass spectrometry data to make high-confidence predictions of protein identities. The underlying computational engine for many of these problems is a massively parallel implementation of a linear programming solver (PCx), which can solve large-scale support vector regression, support vector machine and linear feasibility problems.
Dr. Wagner also is involved in collaborations to perform genome-wide association studies, where his work has concentrated on developing an adequate, rapid data flow infrastructure that includes parallelized genotype calling algorithms.
Education and Training
Dipl. Wi-Ing.: Universitaet Karlsruhe, Germany, 1995.
MS: Operations Research, Cornell University, Ithaca, New York, 1998.
PhD: Operations Research, Cornell University, Ithaca, New York, 2000.
View PubMed Publications
Thompson SD, Sudman M, Ramos PS, Marion MC, Ryan M, Tsoras M, Weiler T, Wagner M, Keddache M, Haas JP, Mueller C, Prahalad S, Bohnsack J, Wise CA, Punaro M, Zhang D, Rosé CD, Comeau ME, Divers J, Glass DN, Langefeld CD. The susceptibility loci juvenile idiopathic arthritis shares with other autoimmune diseases extend to PTPN2, COG6, and ANGPT1. Arthritis Rheum. 2010 Nov;62(11):3265-76.
Freudenberg JM, Sivaganesan S, Wagner M, Medvedovic M. A semi-parametric Bayesian model for unsupervised differential co-expression analysis. BMC Bioinformatics. 2010 May 7;11:234.
Jain R, Wagner M. Kolmogorov-Smirnov scores and intrinsic mass tolerances for peptide mass fingerprinting. J Proteome Res. 2010 Feb 5;9(2):737-42.
Jain A, Velayutham P, Wagner M, Butler DL. Accessing the tissue engineering literature: a new paradigm. Tissue Eng Part A. 2008 Mar;14(3):459-60.