The Move from Cluster Analyses to Neural Networks

5 THE MOVE FROM CLUSTER ANALYSES TO NEURAL NETWORKS 

Due to the mainly qualitative nature of PCA, DFA, and HCA, the role of PyMS in microbiology has been somewhat restricted. For example, using these clustering methods, the use of PyMS to identify microorganisms can be a subjective process because it relies on the interpretation of complex scatter plots and dendrograms. Furthermore the qualitative nature of PCA, DFA, and FICA prevents the application of PyMS to quantitative microbial analysis, while limitations also arise from batch-to-batch variation of PyMS data.80 

Since 1992 a variety of related but much more powerful data-handling strategies have been applied to the supervised analysis of PyMS data. Such methods fall within the framework of chemometrics the discipline concerned with the application of statistical and mathematical methods to chemical data.81-85 These methods seek to relate known spectral inputs to known targets, and the resulting model is then used to predict the target of an unknown input.86 

The first application of ANNs to pyrolysis mass spectra from biological samples was by Goodacre, Kell, and Bianchi.96,97 This study permitted the rapid and exquisitely sensitive assessment of the adulteration of extra-virgin olive oils with various seed oils, a task that previously was labor intensive and difficult. Since this study other laboratories have increasingly sought to apply ANNs to the deconvolution and interpretation of pyrolysis mass spectra, the aim being to expand the application of the PyMS technique from microbial characterisation to the rapid and quantitative analysis of the chemical constituents of microbial and other biological samples. 

Within our laboratory we have shown that using the combination of PyMS and ANNs it is possible to follow the production of indole in a number of strains of E. coli grown on media incorporating various amounts of tryptophan,98 and to estimate the amount of casamino acids in mixtures with glycogen.99 It was also shown that it is possible to quantify the (bio)chemical constituents of complex binary mixtures of proteins and nucleic acids in glycogen, and to measure the concentrations of binary and tertiary mixtures of Bacillus subtilis, Escherichia coli, and Staphylococcus aureus.93,100 

---