Computational approaches for therapeutic target discovery to ameliorate muscle wasting during ageing

In this work I integrated machine learning (ML) approaches, through the development of advanced analytical tools and databases, to characterise the complex miRNA regulatory networks in muscle ageing and identify potential therapeutic targets for sarcopenia. Furthermore, I demonstrated that ML approaches can predict chronological age based on muscle gene expression, providing personalised insights into ageing trajectories. I created a novel software suite, miRKat, to analyse miRNA regulatory networks, integrating multiple data types into a customisable scoring system to prioritise relevant miRNAs. Applied to muscle data, miRKat identified 13 miRNAs of interest, likely mechanistically involved in muscle ageing for further investigation, including known regulators like miR-9 and miR-181, alongside novel candidates such as miR-449. The application of ML, particularly CatBoost, accurately predicted skeletal muscle chronological age from gene expression changes, with an R^2 of 0.96 and an RMSE of 6.90 years, and identified a set of 21 consistently influential genes in muscle ageing signatures. Four of these genes were functionally validated in C. elegans. This approach identified shared and personalised muscle ageing gene signatures and highlighted key pathways involved in muscle. In summary, this research led to the development of miRKat, a computational suite that facilitates the analysis of miRNA regulatory networks, offering a novel perspective on muscle ageing and sarcopenia. The ML approach identified a signature of muscle ageing based on changes in gene expression, highlighting key genes associated with muscle ageing, and network analyses revealed microRNAs targeting these genes. This work extends the current understanding of the mechanisms underlying sarcopenia, identifying candidate biomarkers of muscle ageing and potential therapeutic targets, and providing tools for microRNA research.

Funder

Research Ireland

Publisher

University of Galway

Rights

CC BY-NC-ND

cbn

Collections

University of Galway Theses (PhD Theses)