The 2025 edition of the EMBL Conference ‘Cancer genomics‘ took place in Heidelberg in November, bringing together leading scientists and academic clinicians to further integrate efforts in cancer genomics research.

The discussions covered a range of topics, from cancer genome projects, cancer functional genomics, systems biology, cancer immunogenomics and epigenomics, to the translation and clinical impact of scientific results.

The meeting welcomed 265 on-site participants, two keynote speakers, and 15 invited speakers. In total, 125 posters were presented, offering a broad snapshot of current research. From these, four poster prize winners were selected, and we are pleased to introduce three of them and their work in this post.

Chronic hepatitis b virus infection and xenobiotics exposure interactions at the crossroads of early hepatocellular carcinoma onset

Presenter: Claudia Monge

Authors: Claudia Monge, Agnes Marchio, Stephane Bertani, Juan Pablo Cerapio, Pascal Pineau

Abstract:

In Peru, hepatocellular carcinoma (HCC) unusually affects young patients, including children and adolescents—a pattern not observed elsewhere globally. Most pediatric cases are associated with chronic hepatitis B virus (HBV) infection. Tumors in this group also display strikingly high DNA methylation levels, reminiscent of oncometabolite-driven carcinogenesis. Our recent findings show significant environmental exposure to pesticides in this population, particularly to succinate dehydrogenase inhibitors (SDHi), such as the fungicide penthiopyrad (PTP). However, the potential synergistic effects of chronic HBV infection and SDHi exposure remain unexplored. We investigated transcriptomic alterations induced by PTP in hepatic cell lines with different phenotypes, including HBV-replicating models. PTP exposure triggered stress and inflammatory responses, along with upregulation of mitochondrial genes (MT-ND4, MT-CO2), indicating stimulated mitochondrial function. In low-SDH-expressing cells, inflammatory signaling was amplified, notably pathways mediating T cell and dendritic cell recruitment and neutrophil activation. Enrichment in RIG-I/STING-related pathways suggests a pseudo-viral immune response. In HBV-replicating cells, PTP exacerbated mitochondrial stress and disrupted DNA damage repair mechanisms. Epigenetic profiling revealed a decrease in global DNA hydroxymethylation and H3K27me3 levels—hallmarks of altered chromatin states and potential tumorigenesis. This work uncovers a novel immune-epigenetic interplay linking environmental exposure and chronic infection to an atypical pediatric liver cancer. This mechanism may explain localized cancer patterns and underscores the importance of immune surveillance and environmental triggers in liver tumorigenesis. Our findings highlight the urgent need to address cancer disparities related to environmental injustice and infectious burden in low-resource settings.

Due to the confidentiality of the unpublished data, we cannot share the poster.

Long-read single-cell genome, transcriptome and open chromatin profiling links genotype to phenotypes

Presenter: Luuk Harbers

Authors: Luuk Harbers, Alexandra Pančíková, Ruben Cools, Marios Eftychiou, Margo Aertgeerts, Joris Vande Velde, Heidi Segers, Jan Cools, Jonas Demeulemeester

Abstract:

The rapid advancement of single-cell technologies has revolutionized our understanding of tumour heterogeneity and evolution. However, few multiomics methods are capable of accurately and comprehensively capturing genomic variation at the single-cell level. To address this gap, we developed a novel single-cell multiomics approach integrating high-throughput droplet-based barcoding using 10X Genomics with fragment-length independent nanopore sequencing.
Our method produces full-length barcoded cDNA for precise quantification of known genes, novel transcripts, and complex fusion events, alongside with high-quality DNA data. The latter libraries can be computationally stratified into short fragments from highly accessible regions and longer fragments, enabling simultaneous profiling of chromatin accessibility landscapes and genome-wide variant calling. To achieve maximal genotyping efficiency of variants of interest in all cells, we also propose a targeted approach leveraging these libraries, which allows local sequencing to saturation. To demonstrate its potential, we applied this approach to profile the evolution of pediatric B-cell acute lymphoblastic leukemia (B-ALL) through chemotherapy and CAR T-cell therapy. We reveal clonal dynamics over time and assess the phenotypic effects of somatic variants. Notably, we show how several subclones emerge which escape CAR T-cells through loss of heterozygosity and multiple splice site mutations in the CD19 target. Additionally, we identified transcriptomic and chromatin accessibility shifts linked to immune evasion, offering critical insights into therapy resistance and relapse.
In conclusion, our method delivers simultaneous genomic, epigenomic, and transcriptomic information across thousands of single cells using off-the-shelf kits and reagents, without additional lab equipment. This provides a powerful tool to dissect patterns of somatic evolution, bridging genomic variations to functional insights in unprecedented ways.

View poster

The interplay of mechanism and selection in shaping cancer copy number landscapes

Presenter: Tom Kaufmann

Authors: Tom Kaufmann, Roland Schwarz

Abstract:

Somatic copy-number alterations (SCNAs) are a hallmark of chromosomal instability in cancer, including frequent whole-chromosome, arm-level, and segmental changes. These arise from both mutational processes and selection during tumour evolution. Unlike point mutations, SCNA profiles are ambiguous—multiple event histories can produce the same profile, complicating evolutionary inference. Dependencies between neighbouring segments further obscure selective signals. Prior methods rely on heuristics that lack generalizability, highlighting the need for a principled framework.
We present SPICE (Selection Patterns In somatic Copy-number Events), a method for inferring discrete SCNA events and identifying recurrently selected genomic regions. Using the minimum evolution principle, we analyzed 5,966 TCGA samples across 33 tumour types and inferred distinct copy-number events for them. We show that whole-chromosome and arm-level events are shaped by the density of known tumour suppressors (TSGs) and oncogenes (OGs), while segmental events, though dominated by passenger events (>60%), are locally influenced by the positions of individual genes. Selection loci inferred by tools like GISTIC2 and Biscuit fail to fully account for the observed patterns. To address this, we developed a generative model combining uniform background rates, passenger events, and discrete selection points. This model captures the majority of observed copy number landscapes and recovers most previously reported points of selection, while also identifying an order of magnitude more such loci—revealing that many more genomic regions are subject to copy number–driven selection than previously recognized.

View poster

Poster prize kindly sponsored by Molecular Oncology

The EMBL Conference ‘Cancer genomics‘ took place from 11 – 14 November 2025 at EMBL Heidelberg and virtually.