Jan Hoinka

Combined inference of known and novel mutational signatures with ReDeNovo

ReDeNovo infers both known and novel mutational signatures from cancer genomes, outperforming existing methods and identifying new signatures in real data.

Toward identification of common DNA repair process in mutational signatures

The paper introduces RePrint, a method to infer shared DNA repair mechanisms between mutational signatures, outperforming existing methods and enabling guilt-by-association predictions.

Connecting spatial regions to clinical phenotypes by transferring knowledge from bulk patient data

SpacePhenotyper transfers predictive gene expression-phenotype relationships from bulk data to spatial transcriptomics data to analyze clinical phenotypes in spatial regions of tumors.

RepairSig: Deconvolution of DNA damage and repaircontributions to the mutational landscape of cancer

RepairSig identifies DNA repair deficiency mutational signatures in cancer by modeling interactions between mutagenic processes, replacing multiple signatures with a single representative one in breast cancer.

Co-SELECT reveals sequence non-specific contribution of DNA shape to transcription factor binding in vitro

Co-SELECT identifies DNA shape's non-specific contribution to TF binding in vitro by analyzing HT-SELEX experiments, showing shape features can drive weak binding even without sequence motif similarity.

Accurate sub-population detection and mapping across single cell experiments with PopCorn

PopCorn identifies and maps sub-populations across single-cell experiments with high precision.

AptaBlocks: Accelerating the Design of RNA-based Drug Delivery Systems

AptaBlocks designs RNA sticky bridges to connect aptamers and therapeutic cargoes, enabling efficient RNA-based drug delivery systems.