bims-fragic 2026-04-05 papers

Issue of 2026–04–05
three papers selected by
Laura Mannarino, Humanitas Research

Cancer Epidemiol Biomarkers Prev. 2026 Apr 03.

Early Lung Cancer Detection Using Nucleotide Transition Probabilities in Plasma Cell-Free DNA.

Jinwen Ji, Ruyue Xue, Xu Zhang, Meijia Yang, Lifeng Li, Xiaoran Duan, Wanglong Deng, Rui Yan, Zhihui Xu, Cong Pian, Jie Zhao.

BACKGROUND: Lung cancer is the most lethal malignancy worldwide and urgently requires effective early detection strategies. Current non-invasive approaches based on plasma cell-free DNA (cfDNA) fragmentomics often suffer from limited sensitivity in early-stage patients due to low tumor DNA fractions.
METHODS: We developed a novel computational feature, First-Order Transition Probability (FOTP), to capture nucleotide sequential dependencies within cfDNA fragments. Using low-pass whole genome sequencing data from 1,036 participants, we systematically analyzed cfDNA fragment ends to identify discriminative regions for cancer detection and trained a support vector machine (SVM) model leveraging the FOTP features.
RESULTS: Analysis revealed that the first 10 bp at the 5' end of cfDNA fragments contained the most discriminative information. The SVM model achieved an area under the ROC curve (AUC) of 0.942, with 73.9% sensitivity for stage I and 81.8% for stage II lung cancer at 95% specificity, significantly outperforming existing fragmentomic features. Nucleotide frequency stability and entropy patterns beyond the initial 10 bp supported the biological basis of the approach, reflecting nuclease cleavage biases and chromatin features. The method generalized robustly across independent cohorts and multi-cancer validation sets, showing potential for tissue-of-origin prediction.
CONCLUSIONS: FOTP is a biologically interpretable and highly efficient feature for early cancer detection. It captures key nucleotide dependencies at cfDNA fragment ends, enhancing sensitivity for early-stage lung cancer and other cancers.
IMPACT: This approach offers a scalable and generalizable strategy for early cancer screening.

DOI: https://doi.org/10.1158/1055-9965.EPI-25-2034

Front Genet. 2026 ;17 1794112

Fragmenting the future with FLARE: a comprehensive fragmentomics pipeline based on long-read nanopore sequencing.

Marica Ficorilli, Marta Lucchetta, Deborah Lenoci, Ilenia Rolli, Nicole Farina, Valentina Cristofaro, Lorenzo Giannini, Arianna Ottini, Alberto Deganello, Stefano Cavalieri, Lisa Licitra, Loris De Cecco.

Purpose: Cell-free DNA (cfDNA) fragmentation patterns carry biological information beyond fragment length, revealing nuclease activity, chromatin organization, and tissue of origin. Fragmentomics has emerged as a powerful approach to improve circulating tumor DNA (ctDNA) detection, particularly at low tumor fractions. However, most current methods are designed for short-read sequencing, limiting their applicability to third-generation technologies. Here, we present FLARE (Fragmentation and Long-read Analysis of Regulatory Epigenetics), an integrated fragmentomics pipeline optimized for Oxford Nanopore long-read sequencing.
Methods: FLARE preserves native cfDNA fragment ends and integrates copy number profiling, tumor fraction estimation, sequence-specific end-motif analysis, and methylation-based features to enable comprehensive characterization of cfDNA fragmentation. Plasma cfDNA from six patients with recurrent or metastatic head and neck squamous cell carcinoma (HNSCC) treated with nivolumab was analyzed at baseline (C1D1) and during therapy (C5D1).
Results: Genome-wide copy number analysis revealed recurrent chromosomal alterations consistent with HNSCC biology, with ichorCNA-derived tumor fractions ranging from 0% to 12.8%. Tumor fraction estimates derived from methylation-based fragmentomic features showed concordant trends, providing an independent measure of tumor burden and correlating with clinical response. End-motif analysis based on 5'4-mer frequencies, combined with non-negative matrix factorization (NMF), identified predominant A-end and G-end patterns, consistent with apoptosis-associated nuclease activity.
Conclusion: FLARE provides a robust and scalable framework for fragmentomic analysis using long-read sequencing, enabling simultaneous investigation of structural and sequence-level cfDNA features. This approach demonstrates the technical feasibility of integrated fragmentomic analyses on Nanopore cfDNA data and supports the future integration of native methylation and transcription factor binding site analyses.

Keywords: HNSCC; cfDNA fragmentation; copy number profiling; flare; fragmentation and long-read analysis of regulatory epigenetics; fragmentomics pipeline; head and neck squamous cell carcinoma; long-read nanopore sequencing

DOI: https://doi.org/10.3389/fgene.2026.1794112

Eur J Cancer. 2026 Mar 24. pii: S0959-8049(26)00479-X. [Epub ahead of print]239 116699

Integrating liquid biopsies and artificial intelligence for early cancer detection: A systematic review and meta-analysis.

Panagiotis Filis, Georgios Markozannes, Dimitrios Salgkamis, Nikos Tsiknakis, Ioannis Zerdes, Eirini Pagkalidou, Maria Manou, Nikolaos Filis, Andri Papakonstantinou, Dimitrios Mavroudis, Alexios Matikas, Konstantinos K Tsilidis, Theodoros Foukakis.

INTRODUCTION: The latest generation of liquid biopsies incorporates multi-omic features, including genomics, methylomics, and fragmentomics. Machine learning (ML) approaches have been proposed to synthesize these complex biological data for the development of diagnostic classifiers. This study aims to evaluate the integration of ML with circulating cell-free DNA (cfDNA) analysis for early cancer detection.
METHODS: Medline, Embase, Cochrane, and Web of Science were searched in July 2025. Eligible studies combined ML and cfDNA features to distinguish cancer patients (stages I-III) from non-cancer controls. Summary diagnostic performance metrics and their 95% confidence intervals (CI) were calculated.
RESULTS: The study included 109 articles permitting analyses for lung (n = 34), liver (n = 29), colorectal (n = 28), pancreatic (n = 16), breast (n = 17), esophageal (n = 12), ovarian (n = 13), gastric (n = 9), head and neck (n = 4), and mixed (n = 27) cancer types. Specificity was consistently high across all tumor types and stages (94%-99%). Sensitivity ranged from 72% to 92% for stage I-III, 44-91% for stage I, 71-98% for stage II and 83-99% for stage III. In the pooled study population, neural networks (90%, 95% CI: 81%-95%), random forest (86%, 95% CI: 77%-92%) and heterogeneous ensemble learning (85%, 95% CI: 79%-89%) demonstrated the highest sensitivity. The stratified analysis by classifier feature revealed 86% (95% CI: 80%-90%) sensitivity for fragmentation and 81% (95% CI: 76%-85%) for methylation, with 92%-96% specificity.
CONCLUSION: ML and cfDNA profiling show potential for early cancer detection, with ensemble methods, neural networks and random forests achieving the best overall performance. Fragmentomic features provide the highest sensitivity.

Keywords: Artificial intelligence; CfDNA; Diagnosis; Early-stage cancer; Liquid biopsy

DOI: https://doi.org/10.1016/j.ejca.2026.116699