We assess the potential of quantum computing to accelerate computation of central tasks in genomics, focusing on often-neglected theoretical limitations. We discuss state-of-the-art challenges of quantum search, optimization, and machine learning algorithms. Examining database search with Grover's algorithm, we show that the expected speedup vanishes under realistic assumptions. For combinatorial optimization prevalent in genomics, we discuss the limitations of theoretical complexity in practice and suggest carefully identifying problems genuinely suited for quantum acceleration. Given the competition from excellent classical approximate solvers, quantum computing could offer a speedup in the near future only for a specific subset of hard enough tasks in assembly, gene selection, and inference. These tasks need to be characterized by core optimization problems that are particularly challenging for classical methods while requiring relatively limited variables. We emphasize rigorous empirical validation through runtime scaling analysis to avoid misleading claims of quantum advantage. Finally, we discuss the problem of trainability and data-loading in quantum machine learning. This work advocates for a balanced perspective on quantum computing in genomics, guiding future research toward targeted applications and robust validation.
