bims-fragic 2026-06-14 papers

Issue of 2026–06–14
three papers selected by
Laura Mannarino, Humanitas Research

Cell Rep Med. 2026 Jun 12. pii: S2666-3791(26)00283-1. [Epub ahead of print] 102866

Toward generalizable prediction of cancer signal using a cell-free DNA language model.

Yang Xu, Hua Bao, Daxin Huang, Kun Zhang, Junfei Zhu, Liu Yang, Shengling Fu, Zhili Chang, Jinfeng Zhang, Shuang Chang, Baihan Zhu, Shuyu Wu, Shanshan Yang, Xue Wu, Yang Shao.

Cell-free DNA can be used for early cancer detection, minimal residual disease monitoring, and post-treatment risk stratification. However, current assays are often designed for a single purpose and rely on deep or broad sequencing panels that capture only a small fraction of tumor-derived signals, limiting transferability, increasing cost, and reducing scalability. Fragmentia-AI is an artificial intelligence language model that learns fragment-level sequence patterns in tumor-derived cell-free DNA. Instead of focusing on mutations, it uses the structure of cell-free DNA to detect cancer signals in a partially panel-agnostic manner from ultra-low sequencing input, approximately 0.1%-1% of conventional depth. The model performs well across cancer types and clinical settings, including monitoring after surgery or immunotherapy, and in samples with low variant allele frequencies or no detected mutations. Fragment-level analyses identify shorter fragments and tumor-derived sequence patterns across panels of different sizes and ultra-low-pass whole-genome sequencing in multiple cohorts.

Keywords: artificial intellegence; cell free DNA; early cancer detection; fragmentomics; large language model; minimal residual disease; ultra-low sequencing; whole genome sequencing

DOI: https://doi.org/10.1016/j.xcrm.2026.102866

Nucleic Acids Res. 2026 Jun 08. pii: gkag446. [Epub ahead of print]

DNAviWEB: sequencing-free clinical screening of liquid biopsies.

Anja Hess, Yara Matani, Dominik Seelow, Helene Kretzmer.

Noninvasive cell-free DNA (cfDNA) fragmentation profiles are gaining popularity as diagnostic tools for a wide range of conditions contributing to the global health burden, such as cancer, autoinflammatory disorders, and adverse events in pregnancy or transplantation. Despite their diagnostic value, fragmentomics rely on DNA sequencing, resulting in significant costs and turnaround times, therefore limiting their translation to everyday clinical care. Here, we present DNAviWEB (https://dnavi.sc.hpi.de/), a freely accessible implementation of the DNAvi analysis tool for exploring cfDNA fragmentomics from DNA gel electrophoresis in research settings. DNAviWEB enables instant presequencing analysis of cfDNA while integrating into medical workflows by providing browsable European Genome-phenome Archive and European Liquid Biopsy Society metadata catalogs and delivering rich outputs of patient group-stratified cfDNA statistics, visualizations, and secure storage capacities for clinical data. The DNAviWEB platform opens the exploration, analysis, and database deposition of liquid biopsy DNA profiles to a broad community of clinicians and researchers. DNAviWEB is free and open source.

DOI: https://doi.org/10.1093/nar/gkag446

Genome Med. 2026 Jun 13.

Circulating DNA reveals nucleosome occupancy patterns that are associated with nucleosome-DNA affinity and are affected in cancer.

Marianne Richaud, Ekaterina Pisareva, Paul Burgat, Alain R Thierry, Jacques Colinge.

BACKGROUND: The study of cell-free circulating DNA (cirDNA) fragments (fragmentomics) from liquid biopsies has received increasing attention to detect biomarkers. CirDNA found in blood plasma originates from cells in diverse tissues with a predominance for hematopoietic cells. CirDNA fragments are associated with nucleosomes which protect them against DNAse degradation.
METHODS: The genomic positions of CirDNA fragments originating from different cohorts of healthy individual and cancer patients were obtained from the public database FinaleDB. Various bioinformatic and statistical analyses were conducted on these fragments.
RESULTS: By mapping a large ensemble of well-positioned nucleosomes (WPNs), we found that nucleosome occupancy was associated with histone-DNA affinity, as evidenced by codon usage bias and differences in cirDNA fragment sizes. Moreover, nucleosome occupancy was different in healthy and cancer samples, thus allowing developing a high-performance machine learning approach for cancer detection (specificity and sensitivity > 0.95 for seven cancer types). Cancer influenced nucleosome occupancy in a global manner, although distinct cancer types retained specific features. WPN occupancy at transcription factor binding sites revealed shared, pan-cancer regulation of transcriptional programs involved in hematopoietic cell differentiation and neutrophil biology, the main cirDNA sources.
CONCLUSIONS: This work provides new fundamental insights into cirDNA and DNA sequence using cirDNA as a physical readout. It also bares translational significance by disclosing a new high-performance strategy for cancer detection from liquid biopsies.

Keywords: Cancer biomarker; Circulating DNA; Fragmentomics; Liquid biopsy; Machine learning.; Nucleosome

DOI: https://doi.org/10.1186/s13073-026-01697-9