Computational single-cell investigation of extrachromosomal DNA and replication-timing alterations shaping metastatic cancer

Primary supervisor: Simone Zaccaria, UCL

Secondary supervisors: Benjamin Werner, Queen Mary University of London; Mariam Jamal-Hanjani, UCL

Project

High-level copy-number amplification of oncogenes is a common mechanism of cancer activation and a major source of intratumoural heterogeneity, poor outcomes, and treatment resistance [1]. A key driver of these features is extrachromosomal DNA (ecDNA): circular, centromere-free DNA molecules that carry amplified oncogenes but, unlike chromosomal amplifications, segregate randomly and unequally during cell division [1,2]. This non-Mendelian inheritance allows cancer cells to rapidly change oncogene copy number, generate extreme cell-to-cell heterogeneity, and adapt to therapy far faster than chromosomal mechanisms permit [3]. Despite its recognised clinical importance, how ecDNA evolves during metastatic progression and treatment, and how it reshapes cancer cell states, remains poorly understood.

Single-cell whole-genome sequencing (scWGS) offers a unique opportunity to address this, as cell-to-cell copy-number variability provides a quantifiable readout that can distinguish ecDNA from chromosomal amplifications and reveal its evolutionary dynamics [4]. We recently demonstrated that scWGS can additionally measure replication timing and its alterations, a key epigenetic feature frequently disrupted in cancer, jointly with genetic alterations in the same individual cells [4]. As replication timing reflects the epigenetic state of a cell, this offers a powerful readout of the distinct cancer cell states and the plasticity that underpin treatment resistance, and that may be specifically dictated by ecDNA. However, robust computational methods to detect and characterise ecDNA at single-cell resolution, and to link it to these epigenetic cell states, are still lacking. The goal of this project is to develop and apply computational methods that exploit scWGS to detect ecDNA in individual cells, reconstruct its evolution, and connect it to replication- timing alterations underlying treatment resistance. The project has three aims.

Aim 1. Detecting ecDNA in single cells. We will develop, compare, and benchmark methods to identify ecDNA from scWGS, a challenging problem given the low coverage and high error rates of these data. We are uniquely positioned for rigorous benchmarking using matched samples profiled with deep bulk sequencing and high-throughput, automated DNA FISH [5].

Aim 2. Characterising ecDNA evolution. Using these methods, we will distinguish ecDNA from chromosome-integrated amplifications and study its non-Mendelian dynamics through evolutionary modelling, leveraging the longitudinal and matched pre-/post-therapy structure of our datasets to reveal how ecDNA shapes metastatic progression and resistance.

Aim 3. Linking ecDNA to replication-timing alterations. We will jointly analyse ecDNA and replication timing in the same cells to identify distinct cancer cell states dictated by ecDNA, uncovering mechanisms of resistance and plasticity that extend beyond the amplified oncogenes themselves. The project leverages a unique clinically-annotated, longitudinal scWGS dataset of ~100,000 cells from matched primary and metastatic samples, pre- and post-treatment, across 20 patients within the PEACE autopsy programme [4, 5], expanded through the LUMES project with 300 additional samples, alongside public scWGS datasets of >100,000 breast and ovarian cancer cells with demonstrated ecDNA.

Candidate background

This project would suit candidates with a background in bioinformatics, computer science, mathematics, physics, or a related quantitative discipline, with an interest in algorithm development and in studying cancer evolution. Prior experience in computational method development or the analysis of sequencing data would be an advantage but is not essential.

Potential placements

1. Benjamin Werner, Barts Cancer Institute, Queen Mary University of London
2. Mariam Jamal-Hanjani, Cancer Institute, UCL
3. Nnenna Kanu, Cancer Institute, UCL

References

1. Yan, Xiaowei, Paul Mischel, and Howard Chang. Extrachromosomal DNA in cancer. Nature Reviews Cancer 24.4 (2024): 261-273.
2. Lange, Joshua T., et al. The evolutionary dynamics of extrachromosomal DNA in human cancers. Nature genetics 54.10 (2022): 1527-1533.
3. Bailey, Chris, et al. Origins and impact of extrachromosomal DNA. Nature 635.8037 (2024): 193- 200.
4. Lucas, Olivia, et al. Characterizing the evolutionary dynamics of cancer proliferation in single-cell clones with SPRINTER. Nature Genetics 57.1 (2025): 103-114.
5. Hessey, Sonya, et al. Evolutionary characterization of lung cancer metastasis. Nature (2026): 1-14.