MUSCLE: Multi-platform Unbiased-optimisation of Spectrometry via Closed Loop Experimentation

点击翻译按钮获取中文摘要

Mass spectrometry (MS) is an extremely widely used tool in biology that enables us to measure a wide range of chemicals, including metabolites, peptides and proteins. These measurements are critical for helping us to understand how cells and organisms function at a molecular level. Typically, mass spectrometers are coupled to chromatography. So called liquid chromatography-mass spectrometry (LC-MS) and gas chromatography-mass spectrometry (GC-MS) are extremely common tools in labs in universities and industry. Direct infusion mass spectrometry (DIMS) is also used in some fields due to its measurement rapidity. Regardless of whether a chemist is using LC-MS, GC-MS or DIMS, the development of new analytical methods on a mass spectrometer is extremely time consuming and challenging; e.g., it took one experienced scientist more than a year of effort to develop an optimised LC-MS method for analysing 13 biochemicals (in letters of support). After a method has been published, other scientists will often want to replicate it in their own labs. Yet even this can take considerable time and resources. The primary reason why optimising an MS method is so difficult and time consuming is because the scientist is faced with a very large number of settings for controlling the instrument. Varying all the settings systematically to optimise an analysis is impossible because of the astronomical number of combinations that are possible. So how can we develop these MS methods much more quickly and efficiently? If a solution can be found, labs could develop and implement more MS methods of significantly improved quality, opening up a plethora of novel biological investigations. Also, time savings would translate directly into cost savings, with obvious benefits to universities and industry. Previously we developed computer software that enabled a chemist to optimise automatically their mass spectrometer. We did this for specific LC-MS and GC-MS instruments. Not only was this procedure fully automated, but it greatly improved the analytical method by detecting three times as many biochemicals (revealing new biology), and it only took a few days of automated optimisation to achieve this exciting result. This was published in a leading journal, and read with enthusiasm by the scientific community. Unfortunately, scientists in other labs have not been able to use this software as it was programmed to control only three specific mass spectrometers. Also, it was done in such a manner that reprogramming it for each additional mass spectrometer would be challenging and time consuming. There is now an urgent need for this software to be redeveloped and expanded, so that it can be used to optimise methods on any mass spectrometer in any laboratory. This need, together with the great benefits that would result including considerable time and cost savings, is explained and justified in 12 letters of support that accompany our proposal. These letters are written by scientists in universities, industry and government labs across the world. Our proposal includes two major international companies, GSK and Dionex, as Project Partners. We will develop novel user-friendly software that can control LC-MS, GC-MS and DIMS instruments, which will enable the rapid, robust and fully automated optimisation of MS methods. We will thoroughly test this software on several instruments from several manufacturers. We will also establish this software, and associated 'application control scripts' for controlling a range of mass spectrometers and chromatographs, as a community resource. One way that we will achieve this is by setting up and maintaining a dedicated interactive website for the software. This will include training material (e.g. as a podcast), and the capability for users to upload and share their own application control scripts and to provide feedback. Ultimately this software tool - MUSCLE - promises to facilitate the ever growing use of MS in biology.