The UK population is ageing, as is true across the industrialised world. To solve this grand challenge, science can provide strategies to potentially improve the way humans age. One of this century’s greatest discovery stipulates that ageing follows not a random decline, but rather a genetically controlled process. To study how genes can influence ageing, experimentalists turned to model organisms in which they can measure longevity by studying survival curves.

A major hurdle in such studies is that model organisms can live up to several years, as is the case for mice. Wild-type strains of the nematode C. elegans live just about 2 weeks, a great advantage compared to vertebrate model organisms. To find genes that may influence ageing, experimentalists perform unbiased, functional assays using hundreds of lifespans assays. Although there are several technologies that have now semi-automated the process, a major difficulty is still the time and costs it takes to do experiments and test interventions. Ageing is one of the most complex biological problems at hand, with hundreds of genes and interventions having a small but measurable effect. Computational tools that can help decode complexity into simpler, testable hypotheses can prove to be a valuable asset.

The objective of this study is to apply in silico network tools for the identification of novel genes which may impact the ageing process in C. elegans. Using a time-resolved transcriptomics measurements of three non-reproductive C. elegans strains, I reverse engineered Gene Regulatory Networks (GRN), representative of each strain. I then analysed network properties, and selected candidates to validate experimentally. This strategy tackles one of the main obstacles for ageing research, which is time, but also goes one step further. Not only do GRNs reveal how genes are embedded within a web of molecular interactions, they also point to potential mechanisms by which specific genes may influence ageing.

This study is the first study, to my knowledge, which reconstructs large-scale GRNs of C. elegans, considering over 10,000 genes. Previous work has used perturbation data, but here we used temporally-resolved data of one long-lived and two normal-lived sterile animals, presenting an unbiased network reconstruction that reveals gene interactions based on natural age-dependent gene expression changes.

Finally, we identified and validated several genes which significantly altered the lifespan of the long-lived glp-1(e2144)ts C. elegans, representing a leap forward for in silico tools aim at untangling the complexity of the ageing process.