Fig. 4.
Quantitative correlations within the cation data for PS and GlcCer lipid classes. a, b DTCCSN2 of PS lipids. For the same carbon chain length, PS lipids increase linearly in DTCCSN2 with each loss of unsaturation (a). For the same degree of unsaturation, DTCCSN2 values increase linearly with PS acyl chain length (b). c, d DTCCSN2 of GlcCer lipids. GlcCer species exhibit longer chain lengths than PS, however, the same linear trends are observed for both number of double bonds (c) and for carbon chain length within the same double bond category (d). e A closer inspection of the boxed region highlighted in c demonstrates identification of the initially unknown lipid feature at 802.616 Da, 290.3 Å2 (denoted by a black arrow). Based on the quantitative DTCCSN2 trends, the lipid feature is lower in DTCCSN2 than the predictions for both GlcCer 38:00 + 2Na–H (295.8 Å2, indicated as (i)) and GlcCer 42:06 + H (293.1 Å2, indicated as (ii)). However, the unknown feature aligns well with the DTCCSN2 predicted for GlcCer 40:03 + Na (290.1 Å2, indicated as (iii)), enabling high confidence identification of this lipid. Colors correspond to either summed chain length or degree of unsaturation, whereas shapes correspond to cation type, as specified in the corresponding panel legends. Numerical annotations within symbols correspond either to degree of unsaturation or carbon length, depending on the panel. Error bars represent standard errors for the DTCCSN2 measurements and are all within the size of the markers. Specific lipid identifications, numerical measurement data, and number of measurements per data point (n = 8 or 16 repeat measurements for positive ionization mode data, n = 20, 21, or 28 repeat measurements for negative ionization mode data) are provided in Supplementary Data 1. Information for linear fits including slopes, R-squared values, and number of points per trend line are provided in Supplementary Data 2