Identifying novel omics biomarkers for personalised type 2 diabetes patient profiling and disease prognosis tracking

Current treatment strategies for type 2 diabetes (T2D) are mostly uniform, contributing to inefficient long-term treatment and development of complications. T2D care could be improved using more personalised approaches by further classifying patients based on novel omics biomarkers, including those of disease progression, identified via longitudinal follow-up and genomic profiling. Missingness in phenotype data at one or more time points brings additional limitations for analyses of multiple traits simultaneously.

I will use a ‘reverse regression’ approach that allows analysis of multiple correlated phenotypes jointly to evaluate longitudinal effects, and I will implement an efficient imputation of phenotypes to deal with the missingness issue. I will develop an efficient approach and a software tool for longitudinal high-dimensional omics data analysis that will identify sets of omics biomarkers for personalised longitudinal T2D patient profiles for improved tracking of disease progression.

This project will lead to a large grant application tackling longitudinal multi-omics analyses in large-scale datasets for longitudinal disease prediction.