Techniques for Mass Spectrometry - Cloud Connected Mass Spectrometry Data System

摘要

The 'atomic unit' of data in an liquid chromatography-mass spectrometry(LCMS) experiment is a mass spectrum. While LC apparatus provide compound separation utilities, the detected signal of interest is still the mass spectrum. Historically, because the atomic unit is so information rich it is typically stored as a vendor-specific raw file format Said another way, a collection of mass spectra encapsulating the time domain of the chromatogram is stored in binary format. Post-acquisition processing software typically accesses this data, conducts ETL (extract transform load) operations and runs various signal detection/annotation algorithms. The resulting data is transformed from a raw signal to a biological annotation (e.g molecular formula, peptide sequence etc) Advances in mass spectrometry instrumentation has resulted in a significant increase in data flux and presents a major bioinformatic challenge to traditional desktop applications. Of course, cloud computing strategies can be leveraged to scale algorithms/processing software but the .raw file format limits the ubiquitous application of distributed computing techniques to mass spectrometry specific data. That is, .raw files are the rate-limiting factor of applying massively parallel algorithm techniques such as map-reduce to mass spectrometry data. In other words, the .rawfile is not needed for almost all batch processing algorithms that only require a collection of scans (atomic signal units).