66
line and requirement of a relatively high flow rate for ion current stabilization,
which makes the consumption of a large quantity of lipid samples and thus the
biological material. Moreover, this system is difficult to be automated. These
drawbacks of the delivery system have been now overcome with the introduction of robotic nanoflow ion sources such as NanoMate devices (Wang et al.
2016 and references therein). Further, shotgun lipidomics minimizes the difficulties of lipid analysis from alteration in concentration and ion pairing and
avoids chromatographic anomalies. A constant interaction between lipid species
under constant concentration leads to a constant ratio of ion peak intensities
between lipid species of a class and a constant suppression of lipid species both
within a class and between lipid classes (Vaz et al. 2015; Wang et al. 2016). A
full mass spectrum can be acquired to display the molecular ions of all the species of a lipid class of interest and can be quantified by a direct comparison with
their selected internal standards. Individual lipid species are identified according
to the characteristic fragment ion(s) yielded from the common head group of a
particular lipid class after collision-induced dissociation (Vaz et al. 2015; Wang
et al. 2016). If the characteristic product ion is produced as a result of neutral
loss of a common fragment, then the lipid species are scanned in NLS mode,
while if the characteristic product ion represents the common fragment ion present in the product-ion mass spectra of individual lipid species of the class, then
the lipid species are scanned in PIS mode for monitoring this fragment ion. The
characteristic molecular ions or MS/MS fragmentation data of each lipid class is
summarized in Table 4.1. Shotgun lipidomics is classified into three categories
based on different MS approaches for analysis: tandem mass spectrometrybased, high mass accuracy-based, and multidimensional MS-based shotgun lipidomics (Table 4.2), recently reviewed by Wang et al. (2016).
4.4.3 Lipidomic Data Analysis
A large amount of data is generated in lipidomic experiments and thus requires
preprocessing of data prior to detection and identification of lipid molecules, statistical and biological pathway analysis for reliable outputs, and correct biological
interpretation.
4.4.3.1 Preprocessing of Lipidomic Data
Preprocessing involves the application of methods required to generate a peak table
for each sample containing all detected peaks and their relative concentrations, with
the aim to collect maximum number of true lipid metabolite signals from the data
with minimal number of detected artifacts (Han et al. 2012; Vaz et al. 2015;
P. Kumari
line and requirement of a relatively high flow rate for ion current stabilization,
which makes the consumption of a large quantity of lipid samples and thus the
biological material. Moreover, this system is difficult to be automated. These
drawbacks of the delivery system have been now overcome with the introduction of robotic nanoflow ion sources such as NanoMate devices (Wang et al.
2016 and references therein). Further, shotgun lipidomics minimizes the difficulties of lipid analysis from alteration in concentration and ion pairing and
avoids chromatographic anomalies. A constant interaction between lipid species
under constant concentration leads to a constant ratio of ion peak intensities
between lipid species of a class and a constant suppression of lipid species both
within a class and between lipid classes (Vaz et al. 2015; Wang et al. 2016). A
full mass spectrum can be acquired to display the molecular ions of all the species of a lipid class of interest and can be quantified by a direct comparison with
their selected internal standards. Individual lipid species are identified according
to the characteristic fragment ion(s) yielded from the common head group of a
particular lipid class after collision-induced dissociation (Vaz et al. 2015; Wang
et al. 2016). If the characteristic product ion is produced as a result of neutral
loss of a common fragment, then the lipid species are scanned in NLS mode,
while if the characteristic product ion represents the common fragment ion present in the product-ion mass spectra of individual lipid species of the class, then
the lipid species are scanned in PIS mode for monitoring this fragment ion. The
characteristic molecular ions or MS/MS fragmentation data of each lipid class is
summarized in Table 4.1. Shotgun lipidomics is classified into three categories
based on different MS approaches for analysis: tandem mass spectrometrybased, high mass accuracy-based, and multidimensional MS-based shotgun lipidomics (Table 4.2), recently reviewed by Wang et al. (2016).
4.4.3 Lipidomic Data Analysis
A large amount of data is generated in lipidomic experiments and thus requires
preprocessing of data prior to detection and identification of lipid molecules, statistical and biological pathway analysis for reliable outputs, and correct biological
interpretation.
4.4.3.1 Preprocessing of Lipidomic Data
Preprocessing involves the application of methods required to generate a peak table
for each sample containing all detected peaks and their relative concentrations, with
the aim to collect maximum number of true lipid metabolite signals from the data
with minimal number of detected artifacts (Han et al. 2012; Vaz et al. 2015;
P. Kumari