Comprehensive liquid chromatography-mass spectrometry-based metabolomic analysis of Moringa oleifera seeds

Moringa oleifera seeds which are less explored and nutriment-rich have attracted scientific interest as the seed kernels contain numerous bioactive components with a variety of traditional uses. Besides its medicinal uses, Moringa oleifera biodiesel has shown remarkable potentiality in conducing to the decrease of greenhouse gases and guaranteeing sustainable supply of energy. In this study, the comprehensive analysis of the M. oleifera seeds metabolome was carried out by generating a Molecular Network (MN) from Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS) data to profile the ethyl acetate extract. The dereplication information was then collected by the MN, which then compares the MS/MS spectra of the investigated compounds and groups them into clusters based on their fragmentation route similarities. Therefore, identification of the compounds was conducted based on their full MS and MS/MS spectra obtained in positive ion mode. Through mass spectrometry-based molecular networking a total of 54 metabolites were putatively identified encompassing different classes including coumarins, alkaloids, amino acids, flavonoids, terpenoids, fatty acids, steroids and lipids among others. Thus, the identification highlights that M. oleifera seeds could serve as potential biomarker for new drug discovery and can have a wide variety of applications in food industry. Also, these fatty acids (saturated and unsaturated) suggest that the seed is a good candidate for biodiesel production, since they are fundamental to whether M. oleifera seeds can be used as a biofuel feedstock.

has gained more extensive use in identifying drug metabolite, developing metabolite maps and lending clues to the mechanism of bioactivation (Goulitquer et al., 2012).Using LC-MS/MS based plant metabolomics approach, a few hundreds to thousands of metabolites with high molecular weight (>500 kDa), heat-labile functional groups, chemically unstable functional groups, and high-vapor-point can be detected in a plant extract.It does not require volatilization of the metabolites (Zeki et al., 2020;Piasecka et al., 2019).
LC-MS/MS paired with the computational technique of molecular networking is a cutting-edge data visualization approach that has most notably been used in discovering new drugs from natural sources.The chemical structure of a molecule dictates how it will fragment during the MS/MS procedure (Matt, 2022).Molecular Networking (MN) is a computational strategy used to visualize the structural link between molecules belonging to the same molecular family and interpret complex data arising from MS analysis making it easier to identify unknown metabolites (Messaili et al., 2020).MN is able to identify potential similarities among all MS/MS spectra within the dataset and to propagate annotation to unknown but related molecules (Wang et al., 2016).This approach exploits the assumption that structurally related molecules produce similar fragmentation patterns, and therefore they should be related within a network (Quinn et al., 2017).
In MN, MS/MS data are represented in a graphical form, where each node represents an ion with an associated fragmentation spectrum; the links among the nodes indicate similarities of the spectra.By propagation of the structural information within the network, unknown but structurally related molecules can be highlighted and successful dereplication can be obtained which are useful for metabolite identification (Vincenti et al., 2020;Yang et al., 2013).Metabolite extraction is a critical step prior to metabolomic experiments.The choice of solvents used for extraction is a key factor in determining the metabolites of interest to be extracted, since the main aim of the step is to extract as wide a spectrum of chemical compounds as possible from the sample in consideration (Lu et al., 2017).The ethyl acetate which is a medium polar solvent has been reported to be the best extraction solvent in terms of number of metabolites with large chemical and structural diversity detected by MS (Colnaghi et al., 2007;Beaulieu et al., 2013;Lindow et al., 2014;Di Masi et al., 2022).
Moringa oleifera Lam.belongs to a single genus family Moringaceae; a highly valued plant, distributed in many countries of the tropics and subtropics.It has an impressive range of medicinal uses with high nutritional value.The seed has continued to gain a wider acceptance in various global ethnomedicines for managing several communicable and lifestyle diseases aside its vital nutritional application as emerging food additives.M. oleifera seeds have been shown to elicit a myriad of pharmacological potential and health benefits, including: antimicrobial, anticancer, antidiabetic, antioxidant, antihypertensive, anti-inflammatory and cardioprotective properties.The health benefits of bioactive components in the seeds are promising and demonstrate enough potential to facilitate the development of functional foods (Galuppo et al., 2013;Elsayed et al., 2015;Christian et al., 2022).
Studies on seeds are much accentuated on the purification of water and oil extraction.M. oleifera seeds are used as nature-based solutions for the problem of water purification in developing countries, using them as an alternative to Western methods (El-Haddad et al., 2019).The oil is not only free of toxicant, but it also exhibited high biological value as compared to commercial oil (Saa et al., 2019).Biodiesel produced from M. oleifera seed oil exhibit enhanced oxidative ability, high cloud point and a higher cetane number of approximately 67 which is higher than most biodiesels (Rashid et al., 2008).
However, fewer metabolomics studies have been conducted for large-scale detection of low molecular weight metabolite in M. oleifera seeds which is ideal for incorporation into diets and its metabolite composition contribute to its biological effect.Thus, as liquid chromatography tandem mass spectroscopy (LC-MS/MS) facilitates metabolite identification, quantification and identify patterns in chemical diversity in a complex mixture of molecules, we will gain a better understanding of the properties of the ethyl acetate extract by mapping the chemical profile to the nutritional and pharmacological effects of the Moringa oleifera seeds.
Therefore, this study will employ ultra-high performance liquid chromatography (UHPLC-MS/MS) metabolomics approach to comprehensively profile M. oleifera seeds metabolome to detect the bioactive metabolites present in the ethyl acetate extract.In addition, LC-MS/MS data will be subjected to a molecular networking analysis.

Moringa oleifera seeds preparation
The mature seeds of Moringa oleifera were collected locally from the open market, Bauchi, Bauchi State, Nigeria and authenticated by a taxonomist.The seeds were de-husked manually, air dried and milled into fine powder with the aid of laboratory mortar and pestle.The fine powder was stored at room temperature before extraction.

Solvents and chemicals
Formic acid and acetonitrile of High Performance Liquid Chromatography grade were purchased from Baker (The Netherlands).All other solvents, standards, and chemicals were procured from Sigma Aldrich (St. Louis, MO, USA).

Moringa Oleifera seeds extraction procedure
100g of the powdered seeds was soaked in 300 mL of ethyl acetate (medium polar solvent) for three days at room temperature.
The supernatant was then collected, filtered, and the solvent evaporated using a vacuum rotary evaporator.This step was done twice as according to the methodology described by Jeyaseelan et al. (2012) with little modification.The crude extract was stored at sterile laboratory conditions until further analysis.The ethyl acetate fraction was then used for LC-MS analysis.

Sample preparation for UHPLC-MS/MS analysis
The ethyl acetate extract (2 mg) was dissolved in LCMS-grade methanol (1 mL).Dissolved extract was vortexed for 10 min, centrifuged for 10 min and filtered through a nylon filter (0.22 µm) into a glass vial for LC-MS/MS analysis following method as described by De Oliveira et al. (2017).

Mass spectrometry
High resolution mass spectrometry was carried out using a MicroTOF QIII Bruker Daltonic using an ESI positive ionization with the following settings: capillary voltage, 4500 V; nebulizer pressure, 2.0 bar; drying gas, 8 L/min at 300C.The mass range was at 50-1500 m/z.

Data processing
The accurate mass data of the molecular ions, provided by the TOF analyzer, were processed by Compass Data Analysis software (Bruker Daltonik GmbH).The metabolites characterization was performed using Thermo Xcalibur 2.2.0 (Thermo Fisher Scientific Inc., Waltham, MA, USA) and their comparison was carried out using literature data and standard online databases (freely available), such as PubChem, Human Metabolome Database (HMDB), Chemspider, LIPID MAPS, Metanetx and Swisslipids.In addition, the MS 2 in positive ionization mode and relatively low mass error supported the confirmation of newly identified compounds.

Molecular networking (MN)
The molecular networks based on MS/MS data of M. oleifera seeds were generated using the online workflow Global Natural Products Social Molecular Networking (GNPS) platform (http://gnps.ucsd.edu,accessed on 5th April, 2023) with a registered account.The raw MS data including blank were first converted into mzXML format using MSConvert software downloaded from Proteowizard website (http://proteowizard.sourceforge.net/tools.shtml,accessed on 5th April, 2023) before uploading the data into GNPS.Then, the converted data files were uploaded to GNPS using FileZilla 3.64.0software (https://filezilla-project.org/,accessed on 5th April, 2023).In the GNPS data analysis workflow, sample and blank data were selected as G1 and G2, respectively, with precursor ion mass tolerance set to 0.02 Da and a fragment ion mass tolerance of 0.02 Da.A network was processed with edges that were filtered to have a cosine score above 0.7 and a minimum 6 matched peaks in line with the procedure of Wu et al. (2015).After processing, the spectral networks were imported using the Cytoscape 3.9.1 software, and visualized using a force-directed layout (Institute of Systems Biology, Seattle, WA, USA).

Metabolite profiling of Moringa Oleifera seeds extract Via-UHPLC-ESI MS/MS analysis
The ultra-high performance liquid chromatography (UHPLC) coupled with electrospray ionization (ESI)-micrOTOF-Q III which is a more advanced system known for its high resolution, sensitivity and excellent mass accuracy, was utilized to analyze ethyl acetate extract of M. oleifera seeds.The identification of the compounds was conducted based on their full MS and MS/MS spectra obtained in positive ion mode.All the compounds in the sample were readily ionized in the positive ion mode.The total ion chromatogram (TIC) of the extract is shown in Figure 1, in which a total of 41 chromatographic peaks were annotated.The identities with retention time in minutes (t R ) and fragment ion(s) for each metabolite are presented in Table 1, and the MS/MS spectra with the structures for every molecular ion detected are available in the Supplementary Material (Supplementary Figures S1 and S2).

MS/MS-based molecular networking
Molecular networking (MN) is also known as mass spectral networking.The MN is a graph-based workflow that organizes massive MS datasets by mining spectral similarity between different MS/MS fragmentation patterns, but structurally-related precursor ions.The basic principle underlying MN is to compare the MS/MS spectra of different ions in a sample and to organize those spectra based on similarities.The outcome is a network or graph, in which nodes represent precursor ions and edges represent spectral similarity between the MS/MS spectra of those ions (Nothias et al., 2020).In the present study, the metabolomics mass profile of M. oleifera seeds ethyl acetate extract was analyzed more comprehensively and accurately using the Global Natural Product Social Molecular Networking (GNPS) based on UHPLC-MS/MS analysis data. Figure 6 shows the generated MN with the different clusters in the network, whereby each cluster shares some distinct fragments and fragmentation patterns.The results demonstrated a total of 97 nodes assigned for the parent ions of M. oleifera  1.
seeds.The parent ions spectra in the network was matched with GNPS' spectral libraries and were also identified using different mass spectroscopic databases such as PubChem, HMDB and Chemspider platforms resulting in the annotation of 41 metabolites comprising coumarins, alkaloids, amino acids, flavonoids, terpenoids, fatty acids, and steroids among others Table 1.
Further identification using MN platform on MS/MS data discovered 12 glycerophospholipids that are categorized as (LysoPC, LysoPE, LysoPI, LysoPG, PG) with one glycerolipid as (DG) Table 2. Overall, UHPLC-ESI MS/MS analysis and molecular networking analysis resulted in the tentative identification of 54 compounds in M. oleifera seeds with all of them visualized in different clusters due to their slightly different structures.

Identification of coumarins
Coumarins containing the unique 2H-chromen-2-one motif are secondary metabolites beneficial to human health.They are known for their pharmacological properties such as anti-inflammatory, anticoagulant, antibacterial, antifungal, antiviral, anticancer, antihypertensive, antitubercular, anticonvulsant, antiadipogenic, antihyperglycemic, antioxidant, and neuroprotective properties (Venugopala et al., 2013).Three coumarins were present in the ethyl acetate extract of M. oleifera seeds.Peaks 33 and 41 with t R 15.Another coumarin was putatively identified as 7-Hydroxy-2H-chromen-2-one, (peak 1), also known as umbelliferone at t R 2.15 min.This peak showed a parent ion at m/z 161 [M-H] -ion and characteristics MS/MS fragments at m/z 133 due to loss of CO moiety, m/z 117 corresponding to the loss of CO 2 and m/z 105 produced as a result of loss of 2CO.The proposed mass fragments resulting from the fragmentation of 7-Hydroxy-2Hchromen-2-one is shown in (Figure 2) and is in agreement with previously published result (Zhou et al., 2018).

Identification of amino acids
Amino acids are the fundamental units of proteins, which are also the important components of active peptidases and other bioactive molecules (Liang et al., 2019;Duan et al., 2020).Five amino acid peaks were identified in the UHPLC chromatogram.At t R 4.29 min, peak 3 with m/z 227 was observed to fragment into 209, 199 and 171.These signals result in the loss of a H 2 O molecule, CO and C 4 H 8 respectively.Peak 3 was tentatively identified as Cyclo(L-Leu-trans-4-hydroxy-L-Pro).Peak 4 with m/z 211was identified as Cyclo(L-Leu-L-Pro) by fragment ions m/z 183 and 155 at t R 6.41 min by losses of CO and C 4 H 8 .
Interestingly, peak 11 and 12 with t R 11.79 min and 11.87 min showed similar characteristics MS/MS fragmentation patterns.Peak 11 produced sodiumated molecular ion of m/z 467 and displayed protonated molecular ion of m/z 445 while peak 12 only give

Identification of terpenoids
Terpenoids are known to display a wide range of biological activities which include cancer chemopreventive effects, antimicrobial, antifungal, antiviral, anti-hyperglycemic, anti-inflammatory, anti-parasitic activities and memory enhancers (Kuma et al., 2022).In this study, a total of ten terpenoids were tentatively characterized.and 253 respectively.In addition, the m/z 321 and 319 masses appeared in the same cluster in the MN.Thus, these two peaks suggest they are isomers.

Identification of alkaloids
Alkaloids have shown broad-spectrum antimicrobial activities, and several studies have suggested that these compounds could play an important role in tackling pathogenesis of a variety of infection agents (Cushnie et al., 2008;Casciaro et al., 2020).They are important chemical compounds that serve as a rich reservoir for drug discovery (Lu et al., 2012).UHPLC-MS/MS analysis in positive ionization mode identified five alkaloids, most of which belongs to the isoquinoline type.Dextrorphan with protonated parent ion at m/z 258 was assigned to peak 2 at t R 3.73 min and displaying fragments ions corresponding to successive losses of H 2 O, NH 2 CH 3 and C 4 H 9 at m/z 240, 227 and 201 respectively.Echinulin having an MS parent ion m/z 462 [M+H] + at t R 13.56 min (peak 21) produces ions at m/z 420, 406 and 392 suggesting the losses of C 3 H 6 , C 4 H 8 and C 5 H 10 in a relative manner.
The peak with t R 13.63 min (peak 22) gives a mass peak of m/z 382 [M+H] + is attributed to celecoxib.Some characteristics MS fragments were observed (Figure 3).The first fragment occurred from the loss of one H to produce m/z  (Taheri et al., 2016).

Identification of fatty acid/amides/esters and glycerolipid
These are regarded as important components of lipids necessary for cellular processes in humans and have wide range of commercial applications.As depicted in the UHPLC chromatograms of M. oleifera seeds ethyl acetate extract, one fatty acid, two fatty amide, two fatty

Identification of Organooxygen compounds
Only two organooxygen compounds were identified in the ethyl acetate extract of M. oleifera seeds.Peak 5 was detected at t R 8.47 min with protonated adduct [M+H] + at m/z 205.The MS 2 spectrum was characterized by losses of H 2 O, C 3 H 6 and C 4 H 8 moieties to exhibit fragments at m/z 187, 163 and 149 correspondingly.Peak 32 at t R 14.95 min gave an [M+Na] + ion at m/z 439 and exhibited an [M+H] + ion at m/z 417.The m/z 417 ion yielded product ions at m/z 399, 387 and 373 suggesting losses of H 2 O, CH 2 O and C 2 H 4 O accordingly.

Identification of Flavonoids
Flavonoids possess a number of medicinal benefits, including anticancer, antioxidant, anti-inflammatory, antiviral properties, neuroprotective and cardio-protective effects (Ullah et al., 2020).Three flavonoids were obtained in the UHPLCMS/MS analysis of M. oleifera seeds extract.Peak 6 with t R 9.84 min showed an [M+H] + ion at m/z 303.The m/z 303 ion was subjected to MS/MS analysis and gave product ions at m/z 301, 285 and 126 related to the losses of one H, OH and C 9 H 3 O 3 respectively as shown in Figure 4.By referring to mass spectral libraries, peak 6 was tentatively identified as tricetin.Lysionotin at t R 15.22 min (peak 34) gave an

Identification of benzenoid, dibenzofuran, polyketide and pyridine compounds
These different classes of compounds have various medicinal, agricultural, and industrial applications.Each of benzenoid, dibenzofuran, polyketide and pyridine compounds was produced by M. oleifera seeds ethyl acetate extract.
Benzenoid: Peak 8 at t R 10.49 min with characteristic fragmentation pattern m/z 304, 290 and 278 is attributed to benzyldimethyltetradecylammonium affirming the losses of C 2 H 5 , C 3 H 7 and C 4 H 7 accordingly from the protonated precursor ion m/z 333.
Dibenzofuran: Peak 7 at t R 10.12 min with m/z [M-H] - 369 was putatively identified as didymic acid.The fragment ions at m/z 351, 337 and 325 corresponding to the losses of H 2 O, CH 4 O and CO 2 respectively.Didymic acid is a member of dibenzofuran endowed with antimicrobial activity (Dieu et al., 2012).

Identification of steroids
Plant steroids are unique class of chemical compound that possess many interesting medicinal, pharmaceutical and agrochemical activities like anti-tumor, immunosuppressive, hepatoprotective, antibacterial, plant growth hormone regulator, sex hormone, antihelminthic, cytotoxic and cardiotonic activity (Patel et al., 2015).In the seed extract, three steroids were putatively identified.Peak 14 with m/z [2M+Na] + 1035 and m/z [M+H] + 507 at t R 12.51 min was annotated as esterastin.The MS/MS spectra of esterastin displayed product ion signals at m/z 489 (loss of H 2 O), 434 (loss of C 3 H 5 O 2 ) and 377 (loss of C 6 H 10 O 3 ).
Caudatin (peak 36) at t R 15.77 min produced [M+Na] + ion at m/z 513 displaying [M+H] + ion at m/z 491.The m/z 489 (A) observed from the loss of one H fragment was selected as precursor ion to perform MS 2 analysis from which the m/z 361 (B) was acquired due to the loss of C 7 H 12 O 2 .Three additional fragment ions m/z 343 (C), 325 (D) and 307 (E) were also detected which may be ascribed to the sequential elimination of H 2 0 molecules from (B).To further the investigations, fragmentation patterns of ions B, C and D were chosen as precursor ions in MS 3 analyses to generate ions F, G and H. Ion F at m/z 259 was derived from ion B due to the loss of C 5 H 10 O 2 .Ion G m/z 243 was generated from ion C suggesting the loss of C 5 H 8 O 2 and ion H m/z 283 was produced through elimination of C 2 H 2 O from ion D. Ions B-H were identified as key fragment ions for caudatin (Figure 5).
Another steroid was identified as alphahydroxydeoxycholic acid (peak 38) at t R 16.57 min.The MS/MS spectrum of the peak exhibited fragment ions at m/z 373 and 347 affirming the losses of water and carbon dioxide respectively from the deprotonated precursor ion m/z 391.The fragmentation results agree with previously published data (Chen et al., 2015)

Identification of lipids
Lipids are crucial components of cellular membranes and lipid particles such as lipoproteins, which play many essential roles in cellular functions, including cellular barriers, membrane matrices, signaling, and energy depots (Yang and Han, 2016).Lipids are also valuable energy rich compounds that have the potential to replace conventional fossil fuels through the production of biofuels (Chew et al., 2018).Studies by Takase et al. (2022) indicate that biodiesel made from Moringa oleifera seed oil has a stronger oxidative ability, high cloud point, and a cetane number of around 67, which is higher than most biodiesels.In the current study, the typical nature of the lipid content that makes up the important property of M. oleifera seeds as a suitable candidate for nutrition and biofuel production was revealed through putative annotation of different lipids using MN as shown in Figure 6.The putative identified metabolites matched with the GNPS' spectral libraries and were also identified using different external data bases, namely LIPID MAPS,  glycerol moiety and are defined on the basis of the substituents on the phosphoric acid at the sn-3 position (Farooqui et al., 2000).Glycerophosphocholines (LysoPC) is a monoglycerophospholipid in which the glycerol is esterified with a fatty acid at O-1 position, and  to a glycerol molecule through ester linkages without phosphorylethanolamine moiety.This finding has proven the existence of several monoacylglycerols and diacylglycerols in the lipophilic extracts of M. oleifera seeds obtained from Gas chromatography-mass spectrometry (GC-MS) fatty acid analysis.
Moreover, the seeds harvested in the rainy season, favor the higher content of mono-and diacylglycerides when compared with the seeds harvested during the dry season (Flávia et al., 2022).In another study, fourier transform infrared spectroscopy (FT-IR) and GC-MS analyses were used to characterize the biodiesel in order to investigate the quality and corresponding fatty acid methyl ester (FAMEs) composition in Moringa seed oil, to explore the biodiesel potential (Ruslan et al., 2021).
Molecular networking (MN) facilitates data mining via the clustering of the MS/MS spectra based on fragmentation cosine similarities (Esposito et al., 2017).Figure 7a presents the MS/MS spectrum of LysoPG (15:0/0:0) depicting common fragments found in MS/MS spectrum of LysoPG(13:0/0:0) which appeared in the same cluster.Both shared several fragments such as m/z 229 derived from glycerol moiety and phosphate components and m/z 214 established as cleavage of saturated hydrocarbon chain at its glycerol ester oxygen.
Meanwhile, Figure 7b displayed the represented MS/MS spectrum of LysoPE(0:0/18:1(9Z)) sharing other common fragments with MS/MS spectrum of LysoPE (12:0/0:0) located in the same cluster like m/z 242 forming as glycerol and phosphate backbone free from its ethanolamine moiety and another fragment of unsaturated hydrocarbon chain-forming at m/z 209.
Lastly, some glycerophospholipid which were totally absent in their class cluster due to their slightly different structures were seen in the resulting MN sharing common fragments.The MS/MS spectrums for all metabolites are provided as Supplementary Figure S2).

Conclusion
A novel strategy using UHPLC-QTOF/MS data acquisition combined with the MN was adopted to characterize a large set of metabolites with a wide range of classes in Moringa oleifera seeds.In this research, an efficient exploitation of datasets was employed for automated data treatment and access to dedicated fragmentation databases during MN.The MS/MS-based molecular networking approach, which had never been done on M oleifera seeds has succeeded in the discovery of a  total of 54 metabolites belonging to different bioactive phytochemical classes including coumarins, alkaloids, amino acids, flavonoids, terpenoids, fatty acids, steroids and lipids among others.The results of the study confirms the therapeutic potency of the M. oleifera seeds which could serve as potential biomarker for new drug discovery, and also has a wide application in food industry.
In addition, the presence of saturated and unsaturated fatty acids suggests its benefit for biodiesel production.However, for further studies, fatty acids can be converted into (FAMEs) which could be produced more efficiently by ex-situ transesterification of lipid extracted from M. oleifera seeds.By optimizing the transesterification -reaction involving studying the effect of different reaction conditions, catalysts, and process parameters on the biodiesel production the yield and quality of biodiesel can be improved.

Figure 1 .
Figure 1.Total ion chromatogram (TIC) in positive ionization mode of Moringa oleifera seeds.The number above each peak represents peak numbers, corresponding to the peak numbers in Table1.
14 min and 17.34 min produced [M+Na] + molecular adduct ions of m/z 384 and 419 respectively.Both exhibited [M+H] + ions at m/z 362 and 397.At peak 33, the [M+H] + ion at m/z 362 produced a prominent ion at m/z 344 which was attributed to the loss of a H 2 O molecule.The ion at m/z 344 was further fragmented by losses of CH 4 O and CH 3 COOH leading to the formation of ions at m/z 330, and 302 respectively.Peak 41, the [M+H] + ion at m/z 397 yielded ions at m/z 379, 369 and 355 signaling the losses of H 2 O, CO and C 2 H 2 O in a relative manner.

Figure 3 .
Figure 3. MS/MS spectrum of celecoxib and proposed fragmentation patterns.

Fatty
acid esters: Lauroyl L-carnitine with parent ion m/z [M+H] + 344 at t R 13.75 min was assigned to peak 23.The parent ion dissociated to give fragment ion at m/z 285 as a result of loss of C 3 H 9 N.The characteristics MS/MS fragmentation pattern for peak 31 at t R 14.87 min with fragment ions m/z 409 and 367 corresponding to the losses of H 2 O molecule and CH 3 COOH respectively was attributable to 9-(octanoyloxy) octadecanoic acid having a parent ion m/z [M+H] + 429.Glycerolipid: Peak 17 at t R 13.75 min with m/z 357 [M+H] + was identified as monoeladin based on the fragment ions m/z 339 and 237 due to consecutive losses of H 2 O molecule and C 4 H 8 O 4 in a relative manner.
[M+H] + ion at m/z 345 and an [M+Na] + ion at m/z 367.The [M+H] + ion at m/z 345 revealed fragment ion peaks at m/z 327, 313 and 3011 due to the elimination of H 2 0,CH 4 O and C 2 H 4 O accordingly.Peak 39 with t R 16.68 min produced an [M+H] + ion m/z 611 and an [M+NH 4 ] + ion m/z 628.The [M+H] + ion at m/z 611 displayed fragment ions at m/z 593 and 449 due to loss of H 2 0 molecule and consequent loss of C 6 H 10 O 5 .

Figure 4 .
Figure 4. MS/MS spectrum of tricetin and proposed fragmentation patterns different external data bases, namely LIPID MAPS, HMDB, Metanetx and Swisslipids.The major families identified in the network were the glycerophospholipids), whereas the only glycerolipid identified is the diradylglycerol (DG) as referred to in Table2.The glycerophospholipids are derivatives of sn-glycero-3-phosphoric acid.They contain an O-acyl or O-alkyl or O-alk-1′-enyl residue at the sn-1 position and an O-acyl residue at the sn-2 position of the

Figure 5 .
Figure 5. MS/MS spectrum of caudatin and proposed fragmentation patterns.
linked at position 3 to a phosphocholine.It consists of one chain of stearic acid at the C-1 position.Glycerophosphoethanolamines (LysoPE) are glycerophospholipids where the glycerol is esterified with a fatty acid at O-2 position, and linked at position 3 to a phosphoethanolamine.They are known as lysophospholipid, which refers to any phospholipid that is lacking one of its two O-acyl chains.Glycerophosphoglycerol (LysoPG) are glycerophospholipids, at which only one fatty acid is bonded to the 1-glycerol moiety through an ester linkage.They consist of one chain of saturated fatty acid at the C-1(sn-1) position.Unlike (LysoPG), the (PG) contained two fatty acids bonded to the 1-glycerol moiety through ester linkages.They can have many different combinations offatty acids of varying lengths and saturation attached to the C-1 and C-2 positions.Glycerophosphoinositols (LysoPI) are glycerophospholipids where the glycerol is acylated only at position O-1 with a fatty acid.They can have different combinations of fatty acids of varying lengths and saturation attached at the C-1 (sn-1) or C-2 (sn-2) position.The glycerolipids are a class of lipids containing glycerol to which long-chain hydrocarbons are attached to the hydroxyl groups via carboxylic acid ester linkages (Dowhan et al., 2016).Diradylglycerol (DG), consisting of two fatty acid chains one chain of linolenic acid at the C-1 position and one chain of oleic acid at the C-2 position covalently bonded

Figure 6 .
Figure 6.Full molecular networking of M. oleifera seeds ethyl acetate extract based on tandem mass (MS/MS) spectrometry data in the positive ionization mode identifying 54 various metabolites.Structures shown are representative examples of the lipids identified.Nodes are labeled with parent m/z values and edges are labeled with cosine scores from 0 to 1.

Figure 6 .
Figure 6.Full molecular networking of M. oleifera seeds ethyl acetate extract based on tandem mass (MS/ MS) spectrometry data in the positive ionization mode identifying 54 various metabolites.Structures shown are representative examples of the lipids identified.Nodes are labeled with parent m/ z values and edges are labeled with cosine scores from 0 to 1.

Figure S1 .
Figure S1.MS/MS spectra and structures of various metabolites identified in ethyl acetate extract of M. oleifera seeds in positive ion mode.

Figure S2 .
Figure S2.MS/MS spectra and structures of identified lipids in ethyl acetate extract of M. oleifera seeds in positive ion mode.

Table 1 .
Putative metabolites identified based on LC-MS/MS and MN in ethyl acetate extract of M. Oleifera seeds.

Table 2 .
Putative annotation of lipids identified in ethyl acetate extract of M. Oleifera seeds based on LC-MS/MS and MN.

:
Nonatic acid (peak 10) at t R 10.49 min produced [M+Na] + ion at m/z 225 displaying [M+H] + ion at m/z 203.The ion m/z 203 was subjected to MS 2 analysis yielding product ions m/z 185,157 and 129 due to the losses of H 2 0 molecule, CH 2 O 2 and C 3 H 6 O 2 correspondingly.