Scholar
Ole Winther
Google Scholar ID: 7VAwhzUAAAAJ
Biology, Univ of Copenhagen, Genomic Medicine, Rigshospitalet and Technical University of Denmark
Citations & Impact
Citations
24,242
H-index
60
i10-index
143
Publications
20
Co-authors
97
list available
Publications
Resume (English only)
Research Experience
- Leads research groups at University of Copenhagen and Technical University of Denmark, supervising multiple PhD students
- Current research focuses on applying modern deep learning methodologies to predictive and generative tasks in bioinformatics, NLP, and scientific domains
- Develops inference algorithms and architectures for probabilistic models such as variational autoencoders
- Conducts biological sequence analysis for proteins including signal peptides (SignalP), subcellular localization (DeepLoc), and transmembrane topology (DeepTMHMM)
- Studies regulatory mechanisms at the single-cell gene expression, RNA, and DNA levels
- Long-standing work on latent variable models, including diffusion models, few-shot generation, deep generative models, variational inference, structured mean-field approximations, matrix factorization for collaborative filtering, Gaussian processes, and expectation propagation
- Applies search technologies in health informatics, investigates large language models in medical contexts, and develops end-to-end joint retrieval and generation models
- Creates datasets for training fast surrogate models that replace physical simulations like density functional theory (DFT)
Background
- Professor in High dimensional biological data analysis/Machine learning at Section for Computational and RNA Biology, Department of Biology, University of Copenhagen
- Professor in Data science and complexity at Section for Cognitive Systems, DTU Compute, Technical University of Denmark
- Co-founder and Chief Research Officer (CRO) of raffle.ai, applying modern NLP to enterprise search
- Co-founder and CTO of FindZebra, a search engine for rare diseases
- Head of the ELLIS Unit Copenhagen
- Co-PI of the Machine Learning for Life Science (MLLS) Center
- Research interests include: Bioinformatics (biological sequence analysis, gene expression, genomics), latent variable models (e.g., variational autoencoders, diffusion models) algorithms and architectures, NLP for search and generation, and AI for science
