Bioinformatics Department, Center for Genetic Engineering and Biotechnology, P.O. Box 6162, CP 10600, C. Habana, Cuba.
Anal Chem. 2010 Mar 1;82(5):1759-65. doi: 10.1021/ac9023523.
Fine isotopic structure patterns resolvable by ultrahigh-resolution mass spectrometers are diagnostic of the elemental composition of moderately large compounds. Despite the proven performance of Fourier transforms algorithms to calculate accurate high resolution isotopic distribution, its application to finer ultrahigh resolving power exhibits limited performance. Fast Fourier transforms algorithm requires sampling the relevant range at equally spaced mass values, but ultrahigh resolution mass spectrum displays highly localized complex patterns (peaks) separated in between by relatively large unstructured intervals. Computational efforts consumed on those uninformative intervals are a waste of resources. A fast and memory efficient procedure is introduced in this paper to calculate the isotopic distribution of a single relatively high-mass molecule at ultrahigh resolution by Fourier transforms approaches. The whole isotopic distribution is packed closer to the monoisotopic peak without distorting the actual scale of the peak fine structure. This packing procedure reduced 8 to 32 times the computation resources in comparison to the same calculation performed without packing. The procedure can be readily implemented in existing software.
超高效分辨率质谱仪可分辨的精细同位素结构模式可用于诊断中等大小化合物的元素组成。尽管傅里叶变换算法已被证明可用于计算准确的高分辨率同位素分布，但将其应用于更精细的超高分辨率时，其性能受到限制。快速傅里叶变换算法要求以等间隔的质量值对相关范围进行采样，但超高分辨率质谱显示出高度局部化的复杂模式（峰），它们之间由相对较大的非结构化间隔隔开。在这些无信息间隔上消耗的计算资源是一种浪费。本文介绍了一种快速且节省内存的方法，通过傅里叶变换方法计算超高分辨率下单个相对高分子量分子的同位素分布。整个同位素分布更紧密地集中在单一同位素峰附近，而不会扭曲峰精细结构的实际比例。与未进行包装的相同计算相比，该包装过程将计算资源减少了 8 到 32 倍。该过程可以很容易地在现有的软件中实现。
