Genome-wide association studies (GWAS) provide a correlation of sequence features with disease phenotypes. These associations are often weak and frequently only candidate hits with the best p-values are reported in the literature. However, the selection of a p-value cutoff is rather arbitrary. We propose the use of the "network context" of putative candidate genes within an extended range of p-values. That is, while genes below a certain critical p-value would normally be rejected, we allow their inclusion if they're known to be interacting (genetically or physically) with a better-scoring candidate. One of the best annotated genomes is that of the fruit-fly Drosophila, with most genes being assigned at least a putative function. Our preliminary work suggests that the consideration of tissue-specific functional annotation in Drosophila can be used to establish a network of gene interactions, which in turn helps the interpretation of human disease data. In a previous case, we used functional data about the fly's tracheal system (their breathing organ) to predict interactions of GWAS candidates from a study of lung disease. The PhD project aims to explore the utility of this approach further and gain novel insight into human diseases for which GWAS data as well as equivalent fly data is available (more information upon request). Coding skills are required, as well as a strong interest in human disease.