I can imagine many skills required or helpful for such a position:

• The ability to call/filter high-confidence somatic and germline variants from paired and/or unpaired exome/WGS data,
• annotate both coding and non-coding variants to prioritize recurrent/damaging mutations for functional validation,
• perform experiments to validate hypothesized impact of mutations in cell lines or model systems,
• familiarity with integrated, multi-omics analyses for impactful non-coding variant discovery,
• GWAS,
• software development best practices ranging from proper version control, unit tests, and documentation to project planning, time estimations, and actual programming chops,
• a strong statistics background,
• machine learning skills likely a plus

Most of these assume you're comfortable working with large datasets (100s of samples, TBs of data) in cloud or high-performance computing environments. I'm sure there are many other skills that would prove helpful, but having at least some grasp of most of these will certainly get the ball rolling more quickly if you were to take such a position.