Communication Lower-Bounds for Distributed-Memory Computations for Mass Spectrometry based Omics Data

Abstract

Mass spectrometry based omics data analysis require significant time and resources. To date, few parallel algorithms have been proposed for deducing peptides from mass spectrometry based data. However, these parallel algorithms were designed, and developed when the amount of data that needed to be processed was smaller in scale. In this paper, we prove that the communication bound that is reached by the existing parallel algorithms is Ω(mn + 2r q p ), where m and n are the dimensions of the theoretical database matrix, q and r are dimensions of spectra, and p is the number of processors. We further prove that communication-optimal strategy with fast-memory √ M = mn + 2qr p can achieve Ω( 2mnq p ) but is not achieved by any existing parallel proteomics algorithms till date. To further validate our claim, we performed a metaanalysis of published parallel algorithms, and their performance results. We show that sub-optimal speedups with increasing number of processors is a direct consequence of not achieving the communication lower-bounds proved in this paper. Consequently, we assert that next-generation of provable, and demonstrated superior parallel algorithms are urgently needed for MS based large systems-biology studies especially for metaproteomics, protegenomics, microbiome, and proteomics for non-model organisms.

HiCOPS: High Performance Computing Framework for Tera-Scale Database Search of Mass Spectrometry based Omics Data

“…The model-spectra database can grow exponentially in space (several giga to terabytes) as the post-translational modifications (PTMs) are incorporated in simulation [2], [21]. Therefore, the cost of moving, and managing this data to match with the spectra now exceeds the costs of doing the arithmetic operations in these search engines leading to non-scalable workflows with increasingly larger, and complex data sets [22].…”

“…However, computationally optimal HPC algorithms that minimize both the computational and communications costs for these tasks are still needed. Urgent need for developing methods that exhibit optimal performance is illustrated in our theoretical framework [22], and can potentially lead to large-scale systems biology studies especially for meta-proteomics, proteogenomic, and MS based microbiome or non-model organisms' studies having direct impact on personalized nutrition, microbiome research, and cancer therapeutics.…”