Petroinformatics

Studies on petroleomics have been focused on advanced molecular-level characterization of compounds that could not be analyzed by conventional techniques. The next stage of the development would be more discussions on the information obtained and relationships with the properties and functions. The relationship between molecular composition and bulk properties or functions can be explicitly expressed by petroinformatics, which utilizes statistics, mathematics, and computational visualization technology to interpret or correlate analytical results with bulk properties and experimental data. This provides explicit or implicit information for underlying science and engineering. In this chapter, several examples of petroinformatics are presented. Statistical methods, such as principle component analysis (PCA) for dimensionality reduction in multivariate analysis, and hierarchical clustering analysis (HCA), have been applied to interpret complex petroleum mass spectra obtained by ultrahigh-resolution Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). The mass spectral peaks were statistically analyzed by Spearman’s rank correlation, and by correlation diagrams showing relationships between composition and bulk properties. Additionally, the chapter demonstrates quantitative analyses for petroleum samples by PCA for multivariate analysis and t-tests for univariate analysis. Volcano plots are utilized to visualize the quantitative change or difference between samples in detail. The software platform, which integrates data from many samples obtained from different analytical instruments, is a very important tool to achieve more comprehensive understanding of complex analytes such as crude oils. The learnings from other research fields, such as metabolomics, genomics, and proteomics, are important and valuable for the next steps of petroinformatics development, i. e., standardization of data and retrieval of its metadata information.