Abstract :
[en] The convergence of high-throughput data generation and computational advancements promises a transformative impact across research fields, particularly in human genetics. Se- quencing technologies and multi-omic data, paired with computational power, have brought high expectations for understanding genetic diseases. Yet, despite a vast catalog of genetic information, the molecular basis of most genetic disorders remains unclear, and predictions of variant impact are still insufficiently accurate.
This gap in human genetics stems from two main challenges. First, the standardization and integration of heterogeneous biological data remain complex and inconsistent. While fields like structural biology have successfully curated global datasets, as in the PDB, for high-accuracy protein structure predictions such as AlphaFold, similar efforts in genetics are limited by the diverse and context-dependent nature of biological data. Second, there is an imbalance between resources allocated to sequencing and those invested in experimentally characterizing the functional impacts of genetic variation. To advance predictive accuracy in human genetics, we need systematic, unbiased approaches to experimentally assess the effects of variants and elucidate disease mechanisms.
This thesis explores how model organisms, particularly Saccharomyces cerevisiae, serve as essential platforms for systematic, high-throughput mapping of protein interactions and their perturbations, providing insights into human disease. We propose a scalable frame- work to assess the molecular impacts of missense mutations—single-nucleotide changes that alter one amino acid in a protein—creating a library of variant-containing vectors compat- ible with diverse functional assays across organisms and cell lines. By focusing on protein interaction perturbation and changes in protein abundance, we demonstrate how these complementary approaches can deepen our understanding of the molecular consequences of missense mutations and their roles in disease.
