Cancer Genomics and Diagnostics

The paper identifies ancient matrisomal glycoprotein genes frequently mutated across cancers, showing their role in tissue morphogenesis and neoplastic transformation.

The paper identifies cis-regulatory mutations in cancer that disrupt gene regulatory programs by affecting both transcriptional and post-transcriptional regulation.

This study shows that combining transcriptome data with gene panel features improves drug response prediction in cancer.

MEDICC2 infers cancer phylogenies from copy-number data, accounting for whole-genome doubling and parallel evolution, and performs well on simulated and real datasets.

This paper introduces DF-CAGE, a machine learning method that identifies cancer-druggable genes using multi-omics data from TCGA, finding 465 potential targets and highlighting the importance of CNAs and mutations.

EvAM-Tools provides an R package and web app for building and analyzing evolutionary models of cancer progression and event accumulation.

ROKET uses kernel regression and optimal transport to associate somatic mutations with clinical outcomes, identifying associations in three cancer types.

This study compares genomic profiling of circulating tumor cells (CTCs) using three methods in metastatic colorectal cancer, finding differences in mutational burden and diversity estimates.

SigProfilerExtractor is a tool for de novo mutational signature extraction that outperforms existing methods and identifies four novel signatures, including one linked to tobacco smoking.

The paper analyzes base substitution patterns across species and finds they align with cancer mutational signatures, linking metabolism to DNA damage and mutation generation.

This study compares mutational landscapes in somatic and germline cells across 29 anatomical structures, identifying key mutational signatures and their contributions to different cell types.

This paper identifies mutational signatures caused by formalin fixation and introduces FFPEsig, a method to correct these artefacts in genome sequencing data.

The paper identifies ultra-short single-stranded DNA (ussDNA) in biofluids, showing its presence across species and potential as a DNA biomarker.

The paper introduces OncMTR, a metric identifying genic regions intolerant to germline variation but enriched for cancer-driving mutations, and shows it improves genetic signal detection for hematologic malignancies.

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.

SCONCE2 jointly infers single cell copy number profiles and tumor evolutionary distances, showing higher accuracy than existing methods in simulations and real data.

This study links chromatin accessibility in primary cancers to regional mutation patterns, showing tissue-specific and cancer-type-specific associations with mutagenesis and epigenomes.

This paper uses non-negative matrix factorization to analyze cfDNA fragment length distributions, showing it can infer tumor load, detect cancer, and identify open chromatin-associated signatures.

This study demonstrates the ability to detect cell-of-origin and cancer-specific methylation features from cell-free DNA using Oxford Nanopore sequencing.

The paper develops a method to detect and type cancer using cfDNA sequencing data, achieving 81.62% accuracy in distinguishing malignant from control samples.

CONET infers tumor evolutionary trees and copy number calls from single-cell data using a probabilistic model with MCMC and regularization.

This paper develops a method to estimate tumor mutational burden from RNA-sequencing data without matched-normal samples, achieving high precision and recall in melanoma samples and showing association with patient survival.

This paper introduces PICTograph, a Bayesian model for estimating cancer cell fractions and clone trees from multi-region sequencing data, showing improved accuracy in subclone reconstruction and evolutionary inference.

The paper uses genomic profiling to show that some invasive breast cancers are clonally unrelated to their preceding DCIS, suggesting DCIS may act as both a precursor and independent risk factor.

The paper identifies mutational signatures as markers of drug sensitivity in cancer cells by analyzing drug and genetic screening data.

SigProfilerClusters identifies all types of clustered mutations by calculating sample-dependent thresholds and subclassifies them into established categories.

The paper introduces NEAdriver, a method that combines network analysis and mutation co-occurrence to identify rare cancer drivers in individual tumor genomes, revealing previously undetected driver events.

PhiDsc identifies functionally relevant mutations in protein families by comparing 3D structures and mutation recurrence data, revealing known and novel hotspots in RAS and RHO families.

PersonaDrive identifies and prioritizes patient-specific cancer drivers by integrating genomic data, pathways, and protein connectivity, outperforming existing methods on TCGA and cell-line data.