Hardware acceleration is needed to speed up DNA or protein alignment while keeping the accuracy of the alignment result high enough. The hardware resources used can be effectively optimized by using a new way of run-time reconfiguration, called parameterizable reconfiguration. In this paper, we show how a parameterizable configuration can be used to create a run-time reconfigurable implementation of the Smith-Waterman algorithm. This implementation results in a hardware design that can be implemented on a cheaper FPGA with a performance penalty due to the reconfiguration overhead.