UC Santa Cruz

Abstract Lipids are powerful tools for reconstructing paleoenvironments and geologic processes because of their ubiquity and preservation in the rock record. Traditionally, organic geochemistry studies are targeted analyses of known biomarker compounds with known responses to past environmental conditions or geologic processes. With the recent development of the lipidomics field in the biological sciences, there is an opportunity for non‐targeted analysis to identify new biomarkers. However, lipids in the rock record integrate processes from biologic to geologic (i.e., diagenesis, heating, catagenesis) resulting in more chemically complex matrices than biological samples. As non‐targeted “‐omics” approaches are transferred from biology and applied to sediment, soil, and rock samples to perform geolipidomics, they must be evaluated and potentially adapted for geologic samples. Here we compare four non‐targeted methods of extracting features consisting of a peak area and spectrum from raw gas chromatography mass spectrometry data. We compare targeted data from hydrous pyrolysis experiments of Woodford Shale to non‐targeted features extracted with each method. These comparisons illuminate the pitfalls and/or potential advantages of the different computational approaches when applied to complex matrices. Only one method, MSHub, robustly recreates the targeted data, including a thermal maturity index based upon methylphenanthrenes. Exploration of the MSHub results revealed a wide range of features affected by temperature rise. We illustrate a systematic way to identify new thermal maturity indices in this non‐targeted feature space. Geolipidomics approaches could be broadly applied to generate new hypotheses and novel, robust biomarkers for organic matter sources and Earth‐system processes. Plain Language Summary Lipids are useful for studying ancient environments and geologic processes because they are common and often well‐preserved in sedimentary rocks. Traditional chemical analyses in organic geochemistry focus on specific, known compounds linked to past organisms or conditions. In contrast, biological lipidomics uses a non‐targeted approach that searches for all detectable lipids to discover new biomarkers. Applying this approach to rocks, soils, and sediments is called geolipidomics. Because lipids in rocks come from mixed sources and are altered by burial, diagenesis, and heating, geological samples are often more complex than biological ones, methods must be adapted. In our study, we tested four computational methods for identifying potential chemical fingerprints, or “features,” and compared these non‐targeted results with targeted data from heating experiments on Woodford Shale. This allowed us to evaluate each method for complex rock samples. Only MSHub reliably reproduced the targeted results, including a thermal maturity index used to determine maximum heating temperature. Many MSHub features also changed systematically with increasing temperature. Exploring these results revealed a consistent way to identify new thermal maturity indicators. Overall, geolipidomics provides a powerful path for discovering robust biomarkers that illuminate the origins of organic matter and Earth's geologic and environmental history. Key Points Non‐targeted approaches can characterize samples beyond expected, targeted compounds and can fuel discovery of new biomarkers As non‐targeted “‐omics” approaches are transferred from biology and applied to geologic samples, they must be evaluated and adapted Geolipidomics approaches could be broadly applied to Earth science questions to generate new hypotheses for Earth‐system processes