APEX (Associating Plasma Epigenomics with eXpression)
Year: 2026
Type: software
Abstract: Gene expression defines tumor identity, behavior, and therapeutic response, yet remains difficult to measure noninvasively. Here, we show that circulating cell-free chromatin encodes tumor transcriptional state through structured patterns of histone modifications and DNA fragmentation across regulatory regions. Leveraging this, we develop APEX (Associating Plasma Epigenomic features with eXpression), a computational framework that integrates these signals to infer tumor gene expression at single-gene resolution from plasma. Across diverse cancer types and tumor fractions, APEX accurately recapitulates matched tumor RNA-seq profiles and outperforms existing plasma-based approaches in head-to-head comparisons. Mechanistically, gene-body H3K4me3 and H3K36me3 coverage and fragment-length features emerge as dominant contributors to predictive performance. APEX supports diverse clinical applications, including classification of tumor transcriptional states, discovery of treatment response biomarkers, and nomination of druggable targets across cancer types. This study establishes a chromatin-based encoding of gene expression in plasma and provides a platform for noninvasive measurement of transcriptional programs.
Author: Gunsagar Gulati
Institution: Dana-Farber Cancer Institute
Language: English
Subfield: Cancer Research
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7140436048