Cancer Genomics and Diagnostics

The paper develops a mathematical model to disentangle the sources of clock-like mutations in germline and soma, finding distinct origins for SBS signatures 1 and 5.

This paper introduces MRD-EDGE, a machine learning platform for ultrasensitive ctDNA detection in plasma WGS, showing improved sensitivity and broader applicability for MRD monitoring.

This study uses high-density sampling of human tumours to show that growth occurs uniformly within the volume rather than at the surface.

The paper identifies 48 functional driver genes from copy number alterations in a Drosophila TNBC model and shows their role in increased therapeutic resistance.

The paper develops a standardized data model for cancer genomics to enable interoperable data sharing across countries.

MUSE-XAE extracts mutational signatures from cancer genomes using an explainable Auto-Encoder, achieving accurate extraction and improving tumour-type classification.

The paper develops a method to profile extrachromosomal circular DNA (eccDNA) in plasma and finds higher eccDNA levels in lung adenocarcinoma patients compared to healthy donors.

UMIseq detects circulating tumor DNA in colorectal cancer patients with high sensitivity and specificity, even at low allele frequencies.

The paper introduces MuSE2.0, a tool that accelerates somatic mutation calling for whole-genome and whole-exome sequencing data from heterogeneous tumor samples, achieving significant speed improvements over existing methods.

FinaleMe predicts DNA methylation and tissues-of-origin from plasma whole-genome sequencing by using a non-homogeneous Hidden Markov Model.

The paper investigates the evolution of chromosome 4p loss in basal breast cancer, identifying C4orf19 as a novel gene that suppresses proliferation and is part of a kinase module.

The paper estimates the arrival time of clonal copy number gains in tumors and shows they promote mutational diversification and malignant growth.

This study models how spatially variable cell death in solid tumors affects ctDNA composition and clonal representation, showing potential biases in liquid biopsy profiles.

MuSiCal improves mutational signature analysis by outperforming existing methods in discovery and assignment, revealing new signatures and resolving ambiguities in cancer genomes.

This paper validates a comprehensive genomic profiling assay, K-4CARE, for detecting genetic variants and ctDNA, showing high sensitivity and specificity across multiple cancer types.

This study shows that pre-existing subclones in rectal cancer organoids determine radioresistance, with resistance linked to reduced chromosomal instability.

The paper develops a tool to assess tumour contamination in normal samples using whole-genome sequencing data, finding contamination in various cancer types and highlighting risks to variant calling.

The paper introduces MAGPIE, a computational method to identify driver mutations in cancer pathways by analyzing mutual exclusivity patterns, and validates it on melanoma data.

TCGAnalyzeR provides an integrative web tool for visualizing TCGA data to identify patient cohorts and gene sets for cancer research and clinical decision-making.

The paper identifies two global RMD signatures of mutation rate redistribution in tumors associated with RB1 and TP53 pathway alterations and cell cycle changes.

CRUX is a platform for analyzing cancer genome data, enabling cohort comparisons, biomarker discovery, and visualization with a user-friendly interface and local analysis options.

This paper introduces DETOPT, a method for accurately placing SNVs affected by CNAs in tumor progression trees, showing improved accuracy on simulated data and biologically plausible results on breast cancer metastases.

The paper introduces cancercelllines.org, a resource curating genomic variants in cancer cell lines, providing copy number and SNV data for over 16,000 cell lines.

The paper introduces cfDNA UniFlow, a unified preprocessing pipeline for cell-free DNA data, enabling standardized processing and analysis across different settings.

ZygosityPredictor evaluates the number of gene copies affected by mutations in sequencing data, providing confidence measures for precision oncology applications.

VICTree is a variational inference method for clonal tree reconstruction that uses a Tree-structured Mixture Hidden Markov Model and evaluates on cancer datasets.

This paper develops a simulation-based optimization framework for designing genomic studies focused on clonal evolution, particularly in cancer, to optimize resource allocation and study design.

This study examines genetic, CRISPR-based, and pharmacological dependencies in human cancers by analyzing the dark genome, finding similar dependencies across light and dark genomes and linking dark gene mutations to worse patient outcomes.

This paper identifies mutational processes with high hotspot propensity in tumors and normal tissues, showing that SBS1 and SBS17a/b have distinct genomic factors influencing their hotspot formation.