Niko Beerenwinkel

The paper evaluates perturbation reliability in single-cell data and shows that using reliable perturbations improves model performance and reduces training data needs.

This paper introduces scClone2DR, a framework predicting tumour drug response at the clone level using single-cell DNA/RNA sequencing and drug-screening data, showing improved accuracy in identifying clonal sensitivities and resistance.

This study shows that complex domain adaptation models do not outperform simple gradient boosting models in predicting single-cell anti-cancer drug sensitivity.

This paper introduces a method for inferring phylogenetic trees from single-cell RNA sequencing data by alternating between two tree representations to avoid local optima.

DelSIEVE infers cell phylogeny and variants from scDNA-seq data, improving detection of deletions and evolutionary events compared to existing methods.

This study evaluates the reproducibility of eight read alignment tools using synthetic replicates, finding significant variability in their performance and impact on downstream variant calling.

mt-SCITE reconstructs cell lineages from mitochondrial mutations, outperforming existing methods and revealing subclonal relationships in CD8+ T cells post-vaccination.

FiTree infers fitness landscapes from single-cell tumor mutation trees using Bayesian methods, outperforming existing models and identifying epistatic effects in leukemia data.

OncotreeVIS provides an interactive tool for visualizing mutation tree cohorts and analyzing their properties at the cohort level.

DemoTape is a computational method for demultiplexing targeted single-cell DNA sequencing data, showing improved performance over existing methods for RNA sequencing data.

The paper introduces optimal combinatorial pooling strategies for single-cell sequencing to reduce costs while enabling robust sample identification using genetic profiles and pool compositions.

The paper develops a Bayesian framework to model mutual exclusivity in cancer mutations, finding potential mutual exclusivity between EGFR and IDH1 in glioblastoma multiforme.

VILOCA improves mutation and haplotype calling accuracy for viral sequencing data, especially long-read data, by using quality-aware clustering and Dirichlet process mixture models.

The paper evaluates nine computational tools for identifying SARS-CoV-2 lineages in wastewater sequencing data using simulated samples, finding most tools reliably detect lineages above 5% frequency.

LongSom detects somatic variants in LR scRNA-seq data without normal samples, revealing clonal heterogeneity in ovarian cancer.

CTAT-LR-fusion detects fusion transcripts from long and short read RNA-seq data, improving accuracy in bulk and single cell transcriptomes.

This study introduces metMHN, a model to infer metastatic progression from cross-sectional cancer genomics data, identifying TP53 and EGFR as key mediators and highlighting copy number alterations in post-seeding adaptation.

V-pipe 3.0 is a computational pipeline for analyzing viral genomic data, enabling reproducible and scalable estimation of within-sample genetic diversity.

This paper extends cancer progression models to account for observation bias by modeling how genetic events affect tumor detection, improving the accuracy of inferred causal interactions.

This paper introduces oncotree2vec, a method to embed and cluster tumor mutation trees, capturing genetic relationships for improved tumor subtyping and analysis.

This paper introduces COMPASS, a method to jointly reconstruct mutation and copy number phylogenies from single-cell amplicon sequencing data, showing it detects both clonal and subclonal alterations in leukemia patients.

The paper introduces a gene set enrichment method called pareg that models inter-pathway dependencies using regularized regression, showing improved robustness to noise and ability to recover known pathways in breast cancer data.

This study develops a statistical framework to identify synthetic lethal gene pairs from cancer perturbation screens and validates novel interactions in hepatocellular carcinoma.

SIEVE jointly infers somatic variants and cell phylogeny from scDNA-seq data, showing improved accuracy in both tasks compared to existing methods.

The paper introduces scDEF, a Bayesian model for single-cell RNA-seq data that identifies hierarchical cell states and improves gene signature detection across batches.

The paper introduces bnClustOmics, a Bayesian network mixture model for multi-omics clustering, identifies three HCC subtypes with survival associations, and learns cluster-specific molecular networks.

This paper presents a Bayesian method to call mutations and reconstruct phylogenetic trees from single-cell tumour sequencing data, accounting for mutational loss and recurrence.

This paper introduces structured hierarchical models for probabilistic inference of causal effects from perturbation screening data, showing that incorporating prior knowledge improves parameter inference in noisy datasets.

The paper systematically investigates synthetic operon designs to predict and control protein expression levels in bacterial hosts using machine learning.

scAmpi provides a clinical workflow for scRNA-seq analysis, extracting clinically relevant information and supporting drug treatment identification.