Bile acids are essential steroids with a wide range of biological roles, including the regulation of immunity, nutrient absorption, insulin signaling, appetite, and body temperature. However, due to similarities in their MS/MS spectra, spectral matching with reference MS/MS libraries generally fails to differentiate between isomers. This study introduces a proof-of-concept workflow that uses a mass spectrometry query language filtering tree to distinguish isomeric bile acids in untargeted LC-MS/MS data by leveraging intensity ratios of ions that are close to one another in the MS/MS spectrum. It can be retrospectively applied to existing LC-MS/MS data in data repositories. The filtering tree concept provides the opportunity to annotate and distinguish previously unknown bile acid isomers across LC-MS/MS data sets. To facilitate the ease of applying these filters to LC-MS/MS data sets, we developed a web-based application that simplifies the stepwise filtering tree workflow, removing the need for coding expertise. Here, we apply the multistep filtering application to a representative public data set, which revealed distinct patterns of bile acids associated with different diet types across diverse mammalian species. We further identified the previously uncharacterized bile acid deoxycholyl-N-acetyl-putrescine, which was elevated in carnivores.
Mohanty et al. (Thu,) studied this question.