Publications

Presented at ESMO 2026, this poster highlights findings from a One Biosciences–Institut Curie study published in Cancer Discovery in October 2026. Using single-cell profiling, the study uncovers key mechanisms of chemotherapy resistance in ovarian cancer and shows how tumor composition before treatment can predict patient outcomes. These findings pave the way for earlier identification of patients at risk of relapse and more personalized treatment strategies.

The poster presents early results from the prospective ImmunoMM study conducted with CHU Nantes, showing that single-cell profiling of whole bone marrow can identify both tumor escape mechanisms (e.g., BCMA loss/inactivation) and T-cell dysfunction following T-cell–redirecting therapies. With a 2-week turnaround, the approach integrates target integrity, immune fitness and molecular context to potentially guide treatment sequencing in multiple myeloma.

This poster presents the retrospective SPARK analysis, conducted in collaboration with the Centre Léon Bérard, showing that OneMap™ single-cell transcriptomic profiling can generate clinically interpretable data from archived FFPE samples in patients with advanced solid tumors. Across 16 patients, OneMap produced interpretable multicellular profiles in all samples and identified additional potentially clinically relevant therapeutic hypotheses in 7/16 patients (44%), including tumor-specific target expression and tumor microenvironment features. Presented at ESMO-MAP, these results highlight OneMap’s potential to complement standard molecular profiling and have led to the launch of a prospective SPARK study with Centre Léon Bérard, evaluating its use in real-time Molecular Tumor Board decision-making.

Tumor cellular composition is increasingly recognized as a key driver of treatment response and resistance, but comprehensive profiling remains largely research-based. This study presents a clinically compatible workflow combining single-nucleus RNA sequencing with AI-based cell annotation to profile routine FFPE tumor samples. Validated across 116 specimens and six cancer types, the approach was reproducible, compatible with archival and low-input biopsies, and supports the use of cellular composition as a potential biomarker in precision oncology.

This article introduces a multimodal single-nucleus method to profile gene expression and chromatin marks from frozen tumor biopsies, enabling clinical monitoring of non-genetic tumor heterogeneity. In ovarian cancer patients, it reveals consistent evolution under chemotherapy toward resistant EMT/TNFα programs that are epigenetically primed before treatment, and shows baseline tumor states can predict survival.

This AACR 2025 poster  combines single-nucleus RNA and epigenomic profiling from frozen ovarian cancer biopsies to map tumor cell states and track how they evolve under chemotherapy. It identifies 13 recurring tumor phenotypes, shows treatment reduces cycling cells but enriches TNFα/partial mesenchymal states, and links these patterns to survival—supporting single-nucleus profiling as a scalable way to monitor functional tumor heterogeneity in the clinic.

This AACR 2025 poster uses longitudinal single-nucleus RNA sequencing of metastatic urothelial cancer biopsies to understand why many patients stop responding to immune checkpoint inhibitors. It identifies baseline predictors of response and shows that resistance is linked to reduced antigen presentation and interferon signaling, increased T-cell exhaustion, and a shift toward pro-tumoral (M2-like) macrophages.

MAYA is a computational method for single-cell data that detects distinct pathway activation patterns across cell types, rather than a single average score. It is robust to noise and batch effects, can predict cell types from marker genes, and helps identify shared tumor vulnerabilities across patients