Accurate Multi-Target MS Food Analysis

高度自动化和准确的农药残留分析对于确保人类食品和动物饲料的安全至关重要。高分辨率质谱（HRMS）硬件系统的最新进展使得在单次LC/MS运行中筛选数百种这些目标化合物成为可能，而无需传统LC/MS/MS分析所需的耗时的方法开发或维护。为了使HRMS化合物筛选在实际应用中发挥作用，迫切需要一种高效、准确的软件系统来减少假阳性和假阴性，并进行准确的定量。Cerno Bioscience （Cerno）将通过应用先进的多元计算技术和其开发的独特适用的质谱峰形校准技术来应对这些挑战，从而使完整的质谱信息不仅在质量（m/z）上准确，而且更重要的是在质谱剖面上准确，可以为这些复杂的样品带来承受。除了消除超过质量精度的假阳性外，光谱精度还可以通过放松通常对质量精度的严格要求来减少或消除假阴性。当多元计算技术应用于这种完全校准的质谱剖面数据时，可以检测并消除复杂食品样品中由于不可避免的质谱重叠而产生的更有害的假阴性。作为一个额外的优势，这种完全校准的质谱数据还可以进行质谱域定量，避免容易出错的色谱峰面积集成过程。在该项目的第一阶段，Cerno将：1)开发先进的多元计算技术，以有效处理校准的质谱数据；2)在软件原型中实现对复杂食品基质中数百种化合物进行高精度质谱筛选的计算技术；3)将多达800种标准化合物添加到实际的食品基质中进行研究，并充分建立原理证明。在第二阶段，Cerno将：1)通过与学术和工业伙伴合作，扩展到实际的食品安全分析；2)扩展MS数据系统支持，至少包括三个主要HRMS供应商的仪器；3)构建食品实物HRMS筛选的商业软件产品；4)开始alpha和beta测试，并从3-5个农药残留分析实验室寻求反馈。在二期成功完成后，Cerno将与其学术和行业合作伙伴一起推出这一创新的软件解决方案，并将其提供给政府和行业实验室的食品安全科学家。

DESCRIPTION: Highly automated and accurate analysis of pesticides residuals is essential to ensure the safety of human foods and animal feeds. Recent advancement in High Resolution Mass Spectrometry (HRMS) hardware systems have made it possible to screen for hundreds of these target compounds within a single LC/MS run, without the time-consuming method development or maintenance required of the conventional LC/MS/MS analysis. To make HRMS compound screening practically useful, a highly efficient and accurate software system is urgently needed to minimize both the false positives and the false negatives and to perform accurate quantitation. Cerno Bioscience (Cerno) will meet these challenges by applying advanced multivariate computational techniques and the uniquely suitable MS peak shape calibration technology it has developed so that full MS spectral information, accurate not only in mass (m/z) but more importantly in MS spectral profiles, can be brought to bear for these complex samples. In addition to the elimination of false positives above and beyond mass accuracy, spectral accuracy can reduce or eliminate false negatives by relaxing the often aggressive requirement on mass accuracy. When multivariate computational techniques are applied to such fully calibrated MS profile data, it is feasible to detect and eliminate the far moe harmful false negatives seen in complex food samples due to the inevitable mass spectral overlaps. As an added advantage, such fully calibrated MS data will also enable mass spectral domain quantitation and avoid the error-prone process of chromatographic peak area integration. During Phase I of this project, Cerno will: 1) develop advanced multivariate computational techniques for efficient processing of calibrated mass spectral data; 2) implement the computational techniques in a software prototype for highly accurate mass spectral screening of hundreds of compounds in complex food matrices; 3) investigate with up to 800 standard compounds added into actual food matrices and fully establish the proof-of-principle. In Phase II, Cerno will: 1) expand to actual food safety analysis by working with its academic and industrial partners; 2) expand MS data system support to include at least three major HRMS suppliers' instruments; 3) build a commercial software product for HRMS screening of actual food products; 4) commence alpha and beta testing and seek feedback from 3-5 pesticide residue analysis laboratories. Upon successful completion of Phase II, Cerno will work with its academic and industry partners to launch this innovative software solution and make it available to food safety scientists working in both government and industry laboratories.