MARTIN SLAWSKI

ABOUT ME

I graduated in Statistics at Ludwig-Maximilians-Universitaet Muenchen before joining the Machine Learning Group at Saarland University. My present research focus is sparse recovery and compressed sensing for positive signals, driven by a cooperation with the local Computational Proteomics Group. During this project, we have developed approaches for the automatic deconvolution of isotopic patterns in mass spectrometry. The key ingredients of the ongoing theoretical analysis are elements of convex geometry and random matrix theory.

PROJECTS

Deconvolution of isotopic patterns in computational Proteomics

(in collaboration with the Center of Bioinformatics, Junior Research Group for Computational Proteomics and Protein-Protein Interactions)

In high-resolution protein mass spectrometry, one records signals which are composed of peak patterns, each representing a particular peptide. Identifying these patterns from noisy data is highly non-trivial due to the presence of overlaps (as displayed in the figure above), unknown charge states and irregular deviations from biochemical models.

Sparse Recovery for positive signals

Undetermined linear systems of the form y = Ax in conjunction with sparsity assumptions on x have well been studied in the last years. We work on the special case where A and x are constrained to have nonnegative components only.