Jeroen de Ridder

The paper defines an extended translation initiation sequence (eTIS) and shows that a deep-learning model trained on eTIS features accurately predicts translation initiation, revealing that the Kozak sequence is not optimal and identifying conserved nucleotides that enhance initiation efficiency.

The paper develops a metabolite signature to predict post-weaning diarrhoea in pigs using mass spectrometry and machine learning, but finds no consistent pathway enrichment.

This study uses cfDNA sequencing in newborn foals to detect sepsis-causing bacteria and identifies host-related biomarkers associated with sepsis.

The paper develops a method to directly detect 8-oxo-dG using nanopore sequencing by creating synthetic DNA with context variability and training deep learning models.

This study uses sequence-based deep learning models to identify functional promoter mutations in cancer, finding enrichment in known and novel cancer genes.

This paper introduces NanoRCS, a method using nanopore sequencing and consensus calling to accurately estimate tumor fractions from cfDNA by analyzing SNVs, CNAs, and fragmentomics.

sv-channels improves SV detection precision using deep learning, outperforming GRIDSS while maintaining Manta's recall.

The paper links genes in cancer-susceptibility loci to known somatic driver genes through tissue-specific co-expression networks derived from non-cancer samples.

This paper introduces PARM, a deep learning model trained on MPRA data to predict human promoter activity and identify TF binding preferences, revealing positional and functional differences in regulatory grammar.

This paper develops a validated protocol for using Oxford Nanopore methylation sequencing on FFPE samples to classify brain tumors, enabling point-of-care diagnosis.

The paper performs a GWAS in a case-only design to identify genetic loci influencing age at onset of ischemic stroke and finds a robust association with the APOE locus, suggesting potential survival bias rather than direct causation.

This paper develops a deep learning model for intraoperative pediatric CNS tumor classification using rapid sequencing, achieving accurate results within 60-90 minutes.

This study investigates the mutational consequences of global DNA hypomethylation by analyzing single-cell genetic variation in DNMT1-depleted cells, finding increased chromosomal instability without significant mutational burden or structural changes.

The paper benchmarks and analyzes deep learning models for Nanopore sequencing basecalling, finding that Bonito has the best architecture and that species bias affects performance.

Echtvar compresses variant annotations to enable faster and more efficient SNP and indel analysis.

PolarMorphism identifies pleiotropic SNPs from GWAS summary statistics by converting SNP effect sizes to polar coordinates, revealing shared genetic variants across multiple traits.

This paper develops a machine learning method to classify cancer cell of origin from sparse liquid biopsy data by using data augmentation and integration of multiple genomic features.

MetaboShiny provides an interactive tool for analyzing and annotating direct infusion metabolomics data, demonstrating its use on lung cancer urine samples.

The paper compares two phenotyping approaches for ischemic stroke subtypes and identifies a genetic risk locus for small vessel stroke using the intersect phenotype.

This paper introduces SyNet, a data-driven gene network derived from synergistic gene pairs that improves cancer outcome prediction compared to existing networks.

This paper develops a deep learning classifier using whole genome sequencing data to distinguish cancer types and identify metastatic origins.

The paper introduces a new 3C method, MC-4C, to study multi-way DNA conformations and reveals cooperative and competitive chromatin interactions, including enhancer hubs and loop collisions.

The paper introduces Foresight, a method to interpret Random Forest models in biological datasets by uncovering feature groups and interactions, demonstrating its application on breast cancer gene expression data.

This study uses nanopore sequencing to detect and phase structural variants in patient genomes, identifying novel variants missed by short-read sequencing.