Dr. Lars Hildebrand Diese E-Mail-Adresse ist vor Spambots geschützt! Zur Anzeige muss JavaScript eingeschaltet sein! +49-231-755-6375 Room 2.018 Contact
Dipl.-Inf. Iris Paternoster-Bieker Diese E-Mail-Adresse ist vor Spambots geschützt! Zur Anzeige muss JavaScript eingeschaltet sein! +49-231-755-7719 Room 2.016 Contact

Data analysis in biology and material sciences


The work related to DA-BiMaSc covers two different fields of application: biology and material sciences. Both fields share the same nature of the fundamental data, e. g. spectra, depth profiles, or numerical matrices. Applied methods come from statistics, artificial intelligence, and computational intelligence. Objective of the work is the detection of relationships between the fundamental data and biological or physical properties.

Material sciences: Novel high-tech materials are based on new alloys, new production techniques, or new coating types. One objective is the increase of corrosion resistance, without impairing other technical properties such as paint adhesion, formability and scratch resistance. There is a need to be able to characterise these new types of materials with both fast and
analytically comprehensive techniques. The characterisation is mainly based on the following analytical measuring technologies:

  • Glow Discharge Optical Emission Spectroscopy (GD-OES)
  • Laser ablation spectroscopy (LIBS)
  • X-Ray Fluorescence (XRF) and Diffraction (XRD) spectroscopy
  • Time of Flight Secondary Ion Mass Spectrometry (ToF-SIMS)

The evaluation of the fundamental data is based on statistical methods, as well as methods coming from the field of artificial and computational intelligence. Applied methods are

  • Regression analysis
  • Interactive decision trees
  • Artificial neural networks (ANN)
  • Self organizing maps (SOM)
  • Fuzzy transformation

Biology: Biological sciences such as Genomics, Proteomics and Metabolomics aim at understanding the principles and mechanisms of living cells and organisms on microscopic and molecular level. Experimental data is produced in both qualitative and quantitative manner and contain both already known and yet unknown dependencies or correlations within or between genes and
metabolites. Following technologies are used:

  • Microarray Experiments
  • Nuclear Magnetic Resonance Spectroscopy (NMR)
  • Ion Mobility Spectrometry (IMS)

Statistical and computational methods for the analysis and identification of global and local correlation and dependencies within the data are designed, based on

  • Clustering / Biclustering
  • Evolutionary Algorithms (EA)