A hardware approach to protein identification

At the basis of proteins identification we have a string matching algorithm, which has a computational complexity that scales with the length of both the searched and the reference string. This complexity, as well as the fact that to match a single protein we need multiple search of different string in the whole database, makes the protein identification a computational intensive task taking tens of seconds to complete. When performing this task with General Purpose Processors (GPPs), as it might be in a large scale installation (such as medical or research centers), this long execution time translates into a high energy requirement which greatly impacts the scalability and maintenance cost of the system. This paper illustrates a possible way to exploit Field Programmable Gate Arrays (FPGAs) to implement a string matching algorithm with an higher energy efficiency, up to 6 times better, than a standard GPP; such solution can be a building block for large-scale installations aimed at improving protein identification.