Accelerate multiple sequence alignment by cluster algorithm using reconfigurable hardware

الملخص EN

Sequence alignment of the human genome is a foundation problem in molecular biology; bioinformatician uses this similarity comparison between DNA, RNA, or protein sequences, to find the relationship between organisms or species, and for personal identification.

With the next-generation sequencer, the rate of data generated is exponentially increasing the rate at which it cannot be computationally processed.

Traditional sequence alignment based on PC software's alignment tools requires several hours on state of the art workstations which cannot fulfill the increasing demand for this daily repetitive task.

A hardware-based multiple sequence alignment architecture is described in this manuscript, expresses a comprehensive blueprint of the hardware implementation of small sequence alignment, for pair-wise global alignment technique to achieve high-throughput processing in a far shorter time using reconfigurable hardware, which provides better performance compared to the other platforms.

The experiment was conducted for the simulation study to examine a Parallel Hardware Smith-Waterman algorithm based on Divide and Extended technique (PHSW-DE) on different FPGAs, which changing the curve of Big O notation, leads to GCUPS multiplied by a factor of 10^(M-P), and about 690x faster than used software sequential algorithm.

This work will conclude a solution and provide a reference to further accelerating sequence alignment on an FPGA-based architecture using a parallel algorithm.

As a conclusion for this procedure, the whole human genome multiple sequencing alignment of K-Mer length can be done in less than one hour, to achieve a local hardware sequence alignment in every bioinformatics laboratory.