Postdoctoral and doctoral work, written as case studies.

GATA3 is a major regulator of luminal breast cancer, but its DNA-binding behaviour across the diverse GATA motif landscape, and how that binding translates into regulatory signalling, was not well characterised. I evaluated and benchmarked ChIP-seq tools for alignment, peak detection, and de novo motif analysis, built ATAC-seq and RNA-seq pipelines for differential chromatin-accessibility and expression analysis, and integrated the signals to characterise the regulatory landscape.

Transcripts encoding novel open reading frames, short, previously uncharacterised peptides, were emerging as a potential class of cancer biomarkers and therapeutic targets, but had not been systematically characterised across cancer types. I built a proteogenomic pipeline integrating publicly available RNA-seq and Ribo-seq data across multiple cancer types, and performed functional annotation, structure prediction, disease-mutation mapping, pan-cancer survival analysis, and in-silico drug screening.

Cichlid fishes are a textbook example of rapid speciation, but the genomic substrate that supports that speciation, particularly the role of de novo coding regions and accelerated evolution, was not well mapped using state-of-the-art comparative transcriptomics. I benchmarked read alignment and assembly tools, built an optimised de novo + reference-based comparative-transcriptomics pipeline for two cichlid species, and ran phyloP-based evolutionary rate analysis to identify accelerated regions.

Clinical and experimental observations in skin pigmentation, including how IFN-γ regulates melanosome maturation, needed mechanistic explanations grounded in signalling-network dynamics, in collaboration with three experimental labs. I manually curated and reconstructed signalling and metabolic networks for melanosome transport and melanogenesis from primary literature, built Boolean and ODE-based mathematical models, and ran parameter estimation and sensitivity analyses to identify high-leverage nodes.

Sumoylation is a major post-translational modification system, but how SUMO modification levels respond to perturbations in enzyme activity, and which features of that response are non-obvious, required a model-driven analysis rather than purely descriptive biochemistry. I curated the sumoylation signalling network, developed an ODE-based mathematical model of SUMO conjugation dynamics, and ran steady-state and sensitivity analyses to identify counter-intuitive, system-level features.