ESC

Research

Multi-omics integration

Methods for integrating transcriptomic, mutational, and clinical data to characterise disease heterogeneity, primarily in myeloid neoplasms.

Survival analysis

Heterogeneity-aware multi-omic survival models, including multi-task learning frameworks with frailty components for MDS and AML.

Synthetic data generation

Synthetic generators for omics modalities (DNA methylation, bulk and single-cell RNA-seq, metagenomics), and benchmarking frameworks mapping method performance across the partial information decomposition simplex.

Single-cell genomics

Optimal transport-based analysis of disease progression using CITE-seq multi-modal single-cell data in MDS and AML.

Myeloid neoplasms

Myelodysplastic syndromes (MDS) and acute myeloid leukaemia (AML) are the central disease focus driving the methodological work in multi-omics integration, survival modelling, and longitudinal single-cell analysis.

Graph neural networks

Graph-based representations and message-passing architectures applied to multi-omics integration and patient stratification.