Preparing LC/MS data for cloud and/or parallel image computing
AssigneeInfoclinika, Inc.
InventorAndrey Yur'yevich Bondarenko
Life of this patent
Estimating the term…
Abstract
Functionality is described for data management and querying LC/MS spectrometry data, therefore making it easier to store, retrieve, transfer, and process the mass spectrometry data. The functionality transforms a plurality of raw LC/MS files obtained from a biological experiment into a set of LC/MS images on a common M/Z and RT grid compatible for image processing (e.g., time alignment, peak detection and quantification, differential analysis, etc.). The functionality then spits large LC/MS images into smaller chunks, therefore making easier parallel querying and processing using cloud or high performance computing systems.
Claims21
A method for solving a technical problem of managing increasing volumes of mass spectrometry data in biological experiments by CPU- and memory-intensive image processing so as to facilitate parallel processing, comprising:
- recording liquid chromatography-mass spectrometry (LC/MS) spectra where biological substances are separated by retention time and by mass over charge by different instruments producing raw LC/MS data generated from different formats;
- unifying raw LC/MS data generated from different formats to create LC/MS images;
- chunking the LC/MS images into chunks to facilitate parallel processing, each chunk of each LC/MS image corresponding to the same ranges of retention time and mass over charge of another LC/MS image;
- compressing chunked LC/MS images to facilitate storage and input/output requirements;
- distributing the chunked LC/MS images to a multi-node cloud computing infrastructure;
- providing random access to data within a chunk to minimize number bytes service reads from storage and transferring to a computational node;
- and
- causing parallel processing.
A system for solving a technical problem of managing increasing volumes of mass spectrometry data in biological experiments by CPU- and memory-intensive image processing so as to facilitate parallel processing, comprising:
- format-unifying hardware, the logic structure of which is suitable for:
- recording liquid chromatography-mass spectrometry (LC/MS) spectra where biological substances are separated by retention time and by mass over charge by different instruments producing raw LC/MS data generated from different formats;
- unifying raw LC/MS data generated from different formats to create LC/MS images;
- chunking hardware, the logic structure of which is capable of:
- chunking the LC/MS images into chunks to facilitate parallel processing, each chunk of each LC/MS image corresponding to the same ranges of retention time and mass over charge of another LC/MS image;
- and
- compression hardware, the logic structure of which has the capacity for:
- compressing chunked LC/MS images to facilitate storage and input/output requirements;
- distributing the chunked LC/MS images to a multi-node cloud computing infrastructure;
- providing random access to data within a chunk to minimize number bytes service reads from storage and transferring to a computational node;
- and
- causing parallel processing.
A non-transitory computer-readable medium having stored thereon computer-executable instructions for implementing a method for solving a technical problem of managing increasing volumes of mass spectrometry data in biological experiments by CPU- and memory-intensive image processing so as to facilitate parallel processing, comprising:
- recording liquid chromatography-mass spectrometry (LC/MS) spectra where biological substances are separated by retention time and by mass over charge by different instruments producing raw LC/MS data generated from different formats;
- unifying raw LC/MS data generated from different formats to create LC/MS images;
- chunking the LC/MS images into chunks to facilitate parallel processing, each chunk of each LC/MS image corresponding to the same ranges of retention time and mass over charge of another LC/MS image;
- compressing chunked LC/MS images to facilitate storage and input/output requirements;
- distributing the chunked LC/MS images to a multi-node cloud computing infrastructure;
- providing random access to data within a chunk to minimize number bytes service reads from storage and transferring to a computational node;
- and
- causing parallel processing.
Dependencies are inferred from the claim text.