Reconstrucción del Paisaje de Waddington mediante algoritmos genéticos con análisis big data de datos transcriptómicos

The Waddington epigenetic landscape initially emerged in 1957 as a metaphor for the process of embryonic development into whole organisms. In this metaphor, embryonic development is depicted as a ball rolling down from the highest points (more undifferentiated) to the valleys (more differentiated) following a specific path. This idea has been revisited in recent years, and its construction may lead to a better understanding of differentiation processes.

In this thesis, the construction of the Waddington landscape has been carried out using transcriptomic data available in public repositories. The first step involved developing software for aligning and merging ontologies, aiming to establish a standard for the different cell types within the organism and their differentiation relationships. Using this ontology, a series of algorithms were developed for labeling and downloading samples from the GEO repository. Subsequently, with the labeled samples and the developmental ontology, a genetic algorithm was devised to select a set of genes to construct the Waddington landscape.

Finally, this set of genes was analyzed using Gene Ontology (GO), and compared with genetic signatures of different cell types, yielding significant results for immune system cells and for metabolic and immune processes.