LiPydomics: A Python Package for Comprehensive Prediction of Lipid Collision Cross Sections and Retention Times and Analysis of Ion Mobility-Mass Spectrometry-Based Lipidomics Data

Abstract

Comprehensive profiling of lipid species in a biological sample, or lipidomics, is a valuable approach to elucidating disease pathogenesis and identifying biomarkers. Currently, a typical lipidomics experiment may track hundreds to thousands of individual lipid species. However, drawing biological conclusions requires multiple steps of data processing to enrich significantly altered features and confident identification of these features. Existing solutions for these data analysis challenges (i.e., multivariate statistics and lipid identification) involve performing various steps using different software applications, which imposes a practical limitation and potentially a negative impact on reproducibility. Hydrophilic interaction liquid chromatography-ion mobility-mass spectrometry (HILIC-IM-MS) has shown advantages in separating lipids through orthogonal dimensions. However, there are still gaps in the coverage of lipid classes in the literature. To enable reproducible and efficient analysis of HILIC-IM-MS lipidomics data, we developed an open-source Python package, LiPydomics, which enables performing statistical and multivariate analyses ("stats" module), generating informative plots ("plotting" module), identifying lipid species at different confidence levels ("identification" module), and carrying out all functions using a user-friendly text-based interface ("interactive" module). To support lipid identification, we assembled a comprehensive experimental database of m/z and CCS of 45 lipid classes with 23 classes containing HILIC retention times. Prediction models for CCS and HILIC retention time for 22 and 23 lipid classes, respectively, were trained using the large experimental data set, which enabled the generation of a large predicted lipid database with 145,388 entries. Finally, we demonstrated the utility of the Python package using Staphylococcus aureus strains that are resistant to various antimicrobials.

Figure 1

Schematic representation of the LiPydomics data processing workflow. Input/output files (with corresponding file formats) are depicted in gray. Each cell represents an individual data processing step, and arrows reflect possible workflow sequences. Each cell is color-coded according to the specific module used to perform each step. The consistent and modular API of LiPydomics allows for data processing workflows to be customized to the needs of a particular experiment.

Figure 2

Comparisons of (A, B) ^TWCCS and (C, D) ^TIMSCCS vs ^DTCCS values for lipids in the experimental database. Histograms and CCS-CCS plots provided for the comparisons of the following groups to corresponding overlapping DT values: (A) TW positive mode, (B) TW negative mode, (C) TIMS positive mode, and (D) TIMS negative mode with linear corrections applied. Dotted lines show the linear equation y = x.

Figure 3

Predicted (gold) and measured (purple) lipid CCS values and relative prediction errors for abundant lipid species in the lipid CCS database in (A–C) positive and (D–F) negative ESI modes.

Figure 4

Distributions of predicted (red) and measured (blue) HILIC retention times for major lipid classes (A: DGDG; B: PG; C: PI; D: PE; E: PC; F: LysylPG) spanning the retention time range of the established HILIC method described in the Experimental Section.

Figure 5

Demonstration of linear interpolation retention time calibration using data collected with columns of (A) different lengths or (B) different gradients. Open circles and triangles in (A) represent measured retention times from experiments using 50 and 30 mm columns, respectively, plotted against retention time from the established HILIC method (100 mm column). Open circles and triangles in (B) represent measured retention times from experiments using a faster and slower gradient, respectively, plotted against retention time from the established HILIC method using the same 100 mm column. Solid colored points represent the individual lipids chosen as calibrants, with colors distinguishing between the two experiments. The colored lines reflect the linear interpolation between calibrants that used for converting measured retention times to their reference equivalent.

Figure 6

Illustration of LiPydomics functions by analyzing antibiotic-resistant MRSA strains. (A) PCA projections for parent strain (Par) and strains with resistance to daptomycin, dalbavancin, or vancomycin (Dap2, Dal2, Van4, respectively). (B) PLS-DA projections computed between Par (red) and Dap2 (blue) strains. (C) S-plot showing individual features driving separation between Par (red) and Dap2 (blue) strains. (D-F) Heatmaps of Log₂(fold-change) between Par and Dap2 strains for major bacterial lipid classes. (G-I) Bar plots of individual lipids displaying the most significant differences between Par and Dap2 strains. (J) Number of lipids identified from positive and negative mode data using various combinations of predicted identifiers (m/z, CCS, and/or HILIC retention time).