The WG Expression Profiling focusses on the analysis of expression of biological quantities which have been measured in multiple samples (expression profiles) using high-throughput omics technologies. By employing descriptive and differential analyses with bioinformatics and statistics methods experimental conditions may be characterized and modelled respectively and diagnostic biomarkers may be identified. Therefore methods and algorithms from the following areas are developed, researched and applied:
• descriptive and differential statistical analysis,
• feature selection for the characterization/modelling of experimental conditions and the discovery of biomarker candidates,
• machine learning methods for feature selection and for the modelling of expression-based classifiers,
• annotation of genes and proteins in expression data,
• identification and characterization of known and unknown groups and subgroups using statistical methods as well as machine learning approaches,
• study design and sample size determination,
• survival analysis,
• expression-based network inference and –analysis.
The WG Expression Profiling also provides self-developed software tools (e.g. PAA, SDA) and software workflows as well as a consulting for statistical and computational analyses for our cooperation partners in scientific projects. For the bioinformatics/biostatistics education we offer projects for internships as well as bachelor and master theses.