Computational problems can be classified according to their algorithmic complexity, which is defined based on how the resources needed to solve the problem, e.g. the execution time, scale with the problem size. Many problems in computational biology are computationally infeasible in the sense that the exhaustive search for the optimal solution is prohibitive in practical terms. As a consequence, these problems are tackled through heuristics and approximations aiming to overcome the exceeding computational requirements at the cost of providing suboptimal solutions. The importance of defining the computational complexity of computational biology algorithms is a topic rarely surveyed for broad audiences of bioinformaticians and users of bioinformatics tools. However, recognizing the underlying complexity of any algorithm is essential for understanding their potential and limitations. Thus, the aim of this review is to survey the main algorithmic solutions to intractable problems in computational biology, highlighting the importance of High-Performance Computing in this area.
翻译:计算生物学的许多问题在计算上是不可行的,因为从实际角度来说,彻底寻找最佳解决办法是令人望而却步的,因此,这些问题是通过旨在以提供非最佳解决办法为代价克服超额计算要求的粗略法和近似法来解决的。界定计算生物学算法的复杂性的重要性是很少为生物信息学家和生物信息工具使用者的广泛受众调查的一个专题。然而,认识到任何算法的潜在复杂性对于了解其潜力和局限性至关重要。因此,本审查的目的是调查计算生物学中棘手问题的主要算法解决办法,突出高性计算机在这方面的重要性。