Improving the perormance of the spectral deconvolution stage of the proteomic doscovery process

Mass spectrometry is considered a primary analytical tool for the evaluation of proteins in human body. The protein samples generated using a mass spectrometry instrument are analyzed to detect the protein patterns. The operations performed on mass spectrometric output include spectrum visualization, deconvolution, alignment, normalization, statistical significance tests, and pattern recognition. Bindley Bioscience center in Purdue University’s Discovery Park hosts a Proteomics Discovery Pipeline (PDP). PDP is available as a web-based analysis platform which performs the aforementioned operations. The time taken to generate output in the spectrum deconvolution stage is a major concern. Much of this concern results from the fact that this stage has to deal with large-sized input files that are present in potentially many different input formats. Due to slow operation speed of this stage, overall performance of the PDP is hampered. The research work concentrates on discovering and implementing the ways to increase the performance of the spectral deconvolution stage of PDP. Parallel computing architecture is designed to improve the performance of the software application. Parallel computing improves the performance by splitting the data and task of a given application across different cores or computers. Various existing bioinformatics or other applications have already taken advantage of parallel computing. OpenMP is used as a fine grained technique and Message Passing Interface (MPI) is used as a coarse grained technique in parallelization strategies. This project proposes to implement these techniques in spectral deconvolution stage with the aim that the introduction of these techniques will provide the desired performance gain in the spectral deconvolution stage by leveraging multi-core and multi-node architecture