SpectraX: A Straightforward Tool for Principal Component Analysis-Based Spectral Analysis

The analysis of complex
spectra is an important component of direct/ambient
mass spectrometry (MS) applications such as natural product screening.
Unlike chromatography-based metabolomics or proteomics approaches,
which rely on software and algorithms, the work of spectral screening
is mostly performed manually in the initial stages of research and
relies heavily on the experience of the analyst. As a result, throughput
and spectral screening reliability are problematic when dealing with
large amounts of data. Here, we present SpectraX, a MATLAB-based application,
which can analyze MS spectra and quickly locate m/z features from them. Principal component analysis
(PCA) is used to analyze the data set, and scoring plots are presented
to help in understanding the clustering of data. The algorithm uses
mass to charge (m/z) features to
produce a list of potential natural products.

复杂谱图解析是直接质谱/环境质谱（Mass Spectrometry，MS）应用的重要组成部分，例如天然产物筛选类场景。与依托软件与算法的色谱基代谢组学、蛋白质组学研究方法不同，谱图筛选工作在研究初期大多由人工完成，且高度依赖分析人员的实操经验。因此，在处理大规模数据集时，分析通量与谱图筛选的可靠性均存在显著短板。本文介绍一款基于MATLAB的应用工具SpectraX，其可对质谱谱图进行解析，并快速从中定位质荷比（mass-to-charge ratio, m/z）特征。该工具采用主成分分析（Principal Component Analysis，PCA）对数据集开展分析，并通过得分图辅助用户理解数据的聚类分布情况。该算法通过质荷比特征生成潜在天然产物候选列表。