The group focuses on method development and application of pattern recognition algorithms, both supervised and unsupervised. Tools and software, originating from both own development and external colleagues, are applied to high dimensional data sets, which were often, but not exclusively generated by high throughput and omics technologies (genomics, transcriptomics, different epigenetic readouts). A majority of the data originates from cancer samples or sequencing projects, but the developed software has also been successfully applied to data generated from non-neoplastic biological specimen.
Keywords: Bioinformatics, single cell, multi-omics integration, personalized oncology
Daniel Hübschmann giving a talk on Homologous recombination, BRCAness and the TOP-ART trial @ ITN CONTRA - Computational Cancer Research in 2019:
YAPSA is an R package providing functionality for supervised pattern recognition with a special focus on analysis of mutational signatures:
- Software repository: https://bioconductor.org/packages/release/ bioc/html/YAPSA.html
- Citation: Hübschmann, D., Gu, Z., and Schlesner, M. (2015). YAPSA: Yet Another Package for Signature Analysis. R package version 0.99.10
Bratwurst is a software designed for unsupervised or semi-supervised pattern recognition with non-negative matrix factorization (NMF).
- Software repository: https://github.com/wurst-theke/bratwurst
- Citation: Steinhauser, S.* and Hübschmann, D.* (2016). Bratwurst: preprocessing, wrappers and downstream analysis for NMF. R package version 1.0