Martin Stražar

StriMap models TCR-peptide-HLA interactions using structural and sequence features, achieving state-of-the-art performance and identifying disease-associated peptides in ankylosing spondylitis.

The paper develops a deep learning model to predict peptide-HLA II interactions for post-translationally modified peptides, showing accurate predictions on unseen modifications.

openTSNE is a Python library for t-SNE dimensionality reduction that is faster than existing implementations and supports mapping new data to existing embeddings.

This paper introduces a spatial sequencing method to study host-microbiome interactions in the mouse gut using deep learning for spatial niche detection.

This paper introduces MIST, a deep learning framework for metabolite structure annotation from mass spectra, outperforming existing methods and identifying new metabolites in a disease cohort.

The paper proposes a method to reduce batch effects in single-cell data by embedding new data into a t-SNE visualization of a reference dataset, improving co-clustering of cells from different datasets.