RELIABILITY RANKING AND SCALING IMPROVEMENTS TO THE PROBABILITY BASED MATCHING SYSTEM FOR UNKNOWN MASS-SPECTRA
RELIABILITY RANKING AND SCALING IMPROVEMENTS TO THE PROBABILITY BASED MATCHING SYSTEM FOR UNKNOWN MASS-SPECTRA
复制标题
DOI：
10.1021/ac00281a028
复制
发表时间：
1985-01-01
影响因子：
7.4
通讯作者：
PETERSON, DW
中科院分区：
文献类型：
--
作者：
ATWATER, BL;STAUFFER, DB;PETERSON, DW
本文对五个匹配参数对基于概率匹配（PBM）系统检索正确答案概率的影响作了统计评价。结合在匹配未知光谱时发现的结果值，可以根据预测的匹配可靠性对检索到的参考光谱进行排序。这种排序大大提高了PBM的性能，并且可靠性值特别有助于避免当其光谱实际上不在参考文件中时，最佳匹配光谱代表正确化合物的假设。未知丰度值的二次标度补偿了由仪器变化引起的光谱差异，这是匹配光谱的关键问题。其他改进包括更有效的“标记”技术，以消除虚假的参考峰值。商业GC/MS系统的广泛应用已经证明了这些PBM改进所带来的效率的提高。一个主要的应用是未知化合物的鉴定，这在许多实验室中导致每天产生数百个未知质谱，从而明显需要计算机化的鉴定系统（2-16）。对于代表复杂混合物的样品，不完全的GC分离是不可避免的（17，18）。对于代表多个组分的所得光谱，反向搜索（仅要求参比峰处于未知中）提高了检索性能（4-7）。到目前为止，这种类型的最广泛使用的检索算法似乎是基于概率的
Statistical evaluations of the effects of five matching param-eters on the probability of retrieving a correct answer with the probability based matching (PBM) system have been made. Combining the resulting values found In matching an unknown spectrum makes It possible to rank retrieved reference spectra according to the predicted match reliability. This ranking substantially improves the performance of PBM, and the reliability value Is especially helpful In avoiding the as-sumption that the best matching spectrum represents the correct compound when Its spectrum Is actually notin the reference file. Quadratic scaling of the abundance values of the unknown compensates for spectral differences caused by Instrumental variations, a critical problem In matching refer-ence spectra. Other improvements Include a more effective" flagging" technique to remove spurious reference peaks. Extensive applications with a commercial GC/MS system have demonstrated the Increased effectiveness made possible by these PBM modifications.Thousands of gas chromatograph/mass spectrometers (GC/MS) are now used daily worldwide (J). A major ap-plication is the identification of unknown compounds, which in many laboratories results in the production of hundreds of unknown mass spectra per day, making obvious the need for computerized identification systems (2-16). For samples representing complex mixtures, incomplete GC separation is unavoidable (17, 18)·, for the resulting spectra which represent more than one component reverse searching (only requiring the peaks of the referenceto be in the unknown) improves retrieval performance (4-7). By far themost widely used retrieval algorithm of this type appears to be probability based