INDUCE‑seq: A novel tool for next‑generation risk assessment

O'Dare, Kierney 2025. INDUCE‑seq: A novel tool for next‑generation risk assessment. PhD Thesis, Cardiff University. Item availability restricted.

	PDF - Accepted Post-Print Version Restricted to Repository staff only until 8 December 2026 due to copyright restrictions. Download (34MB)
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (135kB)

Abstract

Within the last decade, there has paradigm shift in the ϐield of toxicology, transitioning awayfrom legacy animal‑based approaches towards high‑throughput in vitro and in silico methods for studying genotoxicity. This has largely been driven by the development of next‑generation risk assessment (NGRA), a human‑relevant, exposure‑led and hypothesis‑driven framework for chemical safety assessment. While traditional methods of genotoxicity testing have en‑ sured public safety, next‑generation sequencing‑based approaches can offer a more nuanced context‑dependent risk assessment by measuring primary DNA damage. The INDUCE‑seq method is an innovative approach for the direct digital mapping of genomic double‑strand breaks (DSBs). This thesis describes the development of INDUCE‑seq for the measurement of chemically‑induced DSBs as to provide a novel tool for NGRA. In Chapter III, INDUCE‑seq was used to measure genomic DSBs in response to acute low‑dose chemical exposure in the physiologically‑relevant TK6 cell line. This illustrates the ability of INDUCE‑seq to detect the genotoxic effects of chemical exposure at considerably lower concentrations compared to legacy cell‑based approaches. Moreover, analysis of the frequency and distribution of DSBs provided insights on the mechanism of action of genotoxic chemicals. Chapter IV describes the application of a powerful unsupervised machine learning (ML) algorithm, volume‑regularised non‑negative matrix factorisation (VRNMF), for systematically exploring the genome‑wide DSB landscape. The development and implementation of VRNMF enabled the extraction of biologically‑relevant latent features from INDUCE‑seq break data. In Chapter V, the VRNMF framework was applied to characterise the breakome and investigate chemically‑induced break patterns. Latent feature extraction by VRNMF identiϐied a total of 12 components, including 10 conserved endogenous components and 2 novel chemical‑ speciϐic components. These components captured distinct break patterns that reϐlect the underlying processes either inϐluenced or induced by chemical exposure. In conclusion, leveraging the direct digital break readout of INDUCE‑seq alongside the VRNMF framework, paves the way for data‑driven evaluation of genotoxicity.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Schools > Medicine
Date of First Compliant Deposit:	8 December 2025
Last Modified:	08 Dec 2025 15:32
URI:	https://orca.cardiff.ac.uk/id/eprint/182964

Actions (repository staff only)

Edit Item