Genome-wide association studies emerged as a powerful tool to detect common variants associated with the disease, but up to this date fail to address missing heritability for many diseases. The main issue in elucidating novel mechanisms of complex diseases is due to the majority of GWAS variants being non-coding, and linking them to the closest gene in genomic space ignoring ample evidence of long-range gene regulation, resulting in gene lists with poor reproducibility between the studies and low rates of experimental validation.Proposed project addresses this major caveat in target gene prediction. First, a recently developed method will be refined by extending the range of datasets used to detect enhancer-promoter associations over different spatiotemporal contexts in humans, and the range of variants for which the predictions are made. Finally, utilising a wealth of information on all human diseases with significant genetic component, this project will focus on epistatic interactions between the loci. The work proposed here will ultimately deepen established international collaborations, with both clinical researchers who have access to pertinent patient data, and later experimental biologists with expertise in validation of non-coding functional elements in relevant model organism for human diseases. Previous experience in working with clinical scientists on diseases like schizophrenia, macular degeneration and non-syndromic neurodevelopmental delay indicated that having means to interpret publicly available data in an automated way, translating findings into experiments, or narrowing down candidates for the validation, and crucially, communicating novel findings in form understandable to clinicians is necessary to speed up translational advancements in this field. I am planning to bring my experience in working in multidisciplinary teams and strongly advocate for collaborations across fields, preferably teaching it to students if opportunity arises.