Recent years have witnessed huge developments in different ‘Omics’ fields, namely genomics, transcriptomics, epigenomics, proteomics, metabolomics and phenomics. The information generated by these ‘Omics’ approaches has enhanced precision and speed to the ongoing breeding programs in developing climate smart and nutrition rich germplasm, which is key for ensuring food security (Parry and Hawkesford, 2012). In recent times, the role of phenomics-based breeding has become evident in improving the crop’s performance, and similarly, genomics has made notable contribution in achieving higher genetic gains (Khush, 2001; Langridge and Fleury, 2011; Wang et al., 2017; Xavier et al., 2017). Nevertheless, the diverse omics platforms have great potential in improving the current understanding of important traits, enabling us to develop new strategies for plant improvement. Among omics approaches, the metabolomics is the most complex and has received inadequate attention in crop science, particularly for trait mapping and plant selections.

Metabolites are indispensable component of plant metabolism owing to their influence on plant biomass and architecture (Turner et al., 2016). In recent years, metabolomics has established itself as one of the major breakthroughs in science, paving the way for accurate metabolite profiling in microbes, plants and animals (Heyman and Dubery, 2016; van Dam and Bouwmeester, 2016; Wuolikainen et al., 2016). Metabolomics has the ability to detect a vast array of metabolites from a single extract, thus allowing speedy and precise analysis of metabolites. In other words, metabolomics offers a comprehensive view of cellular metabolites like small organic compounds, which participate in different cellular events, thus representing the absolute physiological state of a cell. In view of the rapidly advancing metabolomics, the metabolite investigation of mutants and transgenic lines holds potential to understand the metabolic networks and to pinpoint the underlying candidate gene(s) (Fernie, 2003; Yonekura-Sakakibara and Saito, 2006; Hong et al., 2016). Also, the metabolomics helps to resolve gene‘s’ function: how a particular gene impacts upon the metabolic pathway, and uncovers different layers of regulation and interception between linked pathways (Wen et al., 2015), which otherwise is difficult to achieve by conventional assays like microarray (Kusano and Saito, 2012). An integrated approach accommodating inferences from genomics, transcriptomics, proteomics, and metabolomics will allow researchers for cataloging and prioritizing the genes to improve important traits in crop species. Further, above-mentioned omics studies have been further extended to explore the associated regulatory steps such as epigenetic regulation, post-transcriptional and post-translation modification. To this end, the interactome network studies aiming to reveal molecular interactions between biomolecules (nucleic acid, proteins, amino acids, carbohydrates, lipids, etc.) deepen our knowledge about the genotype–phenotype relationship (Vidal et al., 2011; Vadivel, 2015).

Metabolomics is being increasingly used in many crop species irrespective of the availability of transgenic system (Oikawa et al., 2008; Fernie and Schauer, 2009; Daygon and Fitzgerald, 2013; Simó et al., 2014). The metabolomics has the potential to facilitate selection of superior traits and improvement of breeding materials (Zivy et al., 2015). In conjunction with the advances in metabolomics, the availability of whole genome sequence, genome-wide genetic variants and cost-effective genotyping assays opens exciting opportunity to effectively integrate metabolomics in crop breeding programs (Hall et al., 2002; Fernie and Schauer, 2009).

The methods and tools employed in metabolomics study, including the mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy have witnessed substantial improvement. The currently available metabolomics platforms have the capacity to allow large-scale metabolite surveys covering both known as well as unknown metabolites. The deluge of such data, however, makes annotation of metabolites a considerable challenge (Matsuda et al., 2010; Lei et al., 2011). In this context, the ever growing strength of bioinformatics tools coupled with the establishment of metabolomics databases such as the one for model plant Arabidopsis¹, and others for various plant species² have greater implications for metabolite annotation (Tohge and Fernie, 2009; Afendi et al., 2012). A considerable amount of data have resulted from metabolic surveys, which might support plant improvement schemes focusing on the traits of agricultural importance such as yield and stress tolerance. Further, rapid generation of genome scale data by sequencing of DNA and RNA, and by MS quantification of proteins and metabolites necessitates integration of these information in order to devise a holistic way of improving traits of interests (Pandey et al., 2016). Although, most of the current studies are coming in well-established model organisms, such studies may be of common occurrence in other plant species as well. Scientific community currently faces a herculean challenge of dealing with massive multi-omics data for conducting systems-level analyses (Suravajhala et al., 2016). In such scenario, improved statistical and bioinformatics tools will be required to analyze these data sets together for better consolidation, which can eventually be translated for improving plant performance. In this review, we briefly describe about the latest investigations on plant metabolites and the application of metabolomics including metabolic engineering for plant improvement.

The modern metabolomics platforms involve generation of metabolome data using two important techniques, namely NMR and MS (Figure 1). The NMR based metabolite detection relies upon the utilization of magnetic properties of nuclei of atoms under magnetic field. The NMR is a non-destructive method extensively used to identify metabolites with smaller molecular weight (<50 kDa) for diverse applications like metabolite fingerprinting, profiling, metabolic flux and extracting the atomic structural information of compound present in the biological samples (Winning et al., 2009). However, the poor sensitivity of this technique owing to a limited coverage of low-abundance biomarkers poses a major limitation that in turn restricts its extensive use. Unlike NMR, greater sensitivity of MS allows researchers to attain a wide coverage of metabolome data. This led researchers to identify novel metabolic biomarker, and molecules that can facilitate the reconstruction of metabolic pathways and networks. Recently, MS has achieved greater accuracy with the advances in the ionization methods such as atmospheric pressure chemical ionization (APCI), electrospray ionization (ESI) and MALDI-TOF (Issaq et al., 2009). For enhancing the throughput, MS is usually combined with chromatography techniques such as gas chromatography (GC), liquid chromatography (LC), capillary electrophoresis (CE), fourier transform ion-cyclotron resonance (FT-ICR) and field asymmetric waveform ion mobility spectrometry (FAIMS). Notwithstanding the low sensitivity and large sample requirement of NMR, its capacity of identifying physical properties of ligands, binding sites on protein, uncovering structures of protein ligand complexes and direct binding of target protein retains its use over MS.

The GC-MS platform is widely used for non-targeted analysis (Dutta et al., 2012). GC-MS approach involves derivatization of samples which makes the compounds volatile; however, this leaves underivatized compounds (except hydrocarbon) unnoticed during analysis. Introduction of GC X GC-TOF-MS has notably improved the separation of co-eluting peaks (deconvoluted peak) and also facilitated higher sample throughput (Ralston-Hooper et al., 2008). LC-MS mostly uses ESI and APCI; it has been widely used for targeted and non-targeted approach to detect both primary and secondary metabolites of higher mass i.e., <1500 Da (Turner et al., 2016). Additionally, the combination of UPLC with QTOF-MS has increased the peak resolution, mass accuracy and rapid identification of hundreds of metabolites in a short span of time (Chawla and Ranjan, 2016). In addition to these platforms, CE-MS provides high-resolution separation of different groups of analytes (charged, neutral, polar and hydrophobic) in both targeted and untargeted metabolomics studies (Ramautar and de Jong, 2014). FT-ICR-MS driven by high-resolution mass analysis facilitates detection of large-scale metabolite species with high accuracy (Brown et al., 2005), which could also be combined with separation techniques in order to resolve “very complex matrices” (Schrader and Klein, 2004) and to tackle other issues including ion separation (Lopes et al., 2017). Additionally, FAIMS or differential mobility spectrometry (DMS) an ion mobility based electrophoretic technique coupled with MS. The FAIMS technology was used for the detection of biological samples like volatile compounds formed during bacterial growth (Zrodnikov and Davis, 2012).

Exhaustive data set generated from above high throughput tools are processed through data processing platforms like MET-COFEA, Met-Align, ChromaTOF, MET-XAlign, etc., (Pegasus, 2007; Lommen et al., 2012; Kessler et al., 2014; Zhang et al., 2014, 2015; Ma et al., 2016; Misra and van der Hooft, 2016; de Souza et al., 2017). This basically includes baseline correction, alignment, separation of co-eluting peaks (deconvolution), normalization, etc. (Figure 1) prior to identification of compounds. Metabolome databases like METLIN, NIST, GOLM etc., can be used for identification of metabolites (Johnson and Lange, 2015). Further, the identified metabolites data are subject to statistical analysis such as correlation map, principal component analysis (PCA), partial least squares (PLS), K-means clustering, boxplot, heatmap, reconstructing metabolic pathways etc., by using web tool and softwares such as MetaboAnalyst, Cytoscape, Statistical analysis tool etc., (Tsugawa et al., 2015; Xie et al., 2015). These analyses are useful to monitor and identify metabolic markers associated with several agronomic traits.

Metabolomics studies have provided greater insights in fruit biology specially related to ripening and quality. Tomato (Solanum lycopersicum) is a rich source of carotenoids, anti-oxidants and flavonoids (Tohge and Fernie, 2015). Metabolite segregation pattern of 50 tomato cultivars showed close agreement with segregation of fruit’s size (Moco et al., 2008). Metabolome is useful to dissect the ripening event by plotting a correlation with fruit transcriptome (Carrari et al., 2006; Osorio et al., 2011). Metabolome can be used to elucidate diverse and differential biochemical pathways exist in the fruits of tomato ILs and Ecotypes (Toubiana et al., 2012; Upadhyaya et al., 2017), and the ancestral species through genome wide metabolic survey (Perez-Fons et al., 2014).

Apple (Malus spp.) contains beneficial nutrients in the peel and flesh, including antioxidants that reduce the risk of chronic diseases such as asthma, cancer, cardiovascular disease, and diabetes (Boyer and Liu, 2004). The metabolite contents of the apple fruits are used to differentiate commercially important cultivars (Cuthbertson et al., 2012). For example, the cultivar ‘Golden Delicious’ contains a high load of myo-inositol, sugars and succinic acid; whereas, the cultivars ‘Red Delicious’ and ‘Fuji’ show relatively higher abundance of triterpene/sterols, flavonoids, phenolic acids, stearic acid, anthocyanins, and carbohydrates. The fruit peel extract of ‘Fuji’ contained high levels of carbohydrate including glucose and sorbitol, and was significantly differentiated from ‘Red Delicious’ and ‘Granny Smith’ which contain high levels of unsaturated fatty acids (oleic and linoleic acid). Spatial distribution of sugars and organic acids between fruit layers has been elucidated in a recent study on metabolic profiling of apple fruit (Cebulj et al., 2017). The browning of apple fruits during storage renders them unmarketable, thus exerting an adverse impact on the apple industry. The metabolomics study on stored apple fruits showed a difference in the level of primary metabolites with different time duration (Hatoum et al., 2014). The increased levels of mannose and xylose during post-harvest indicated a breakdown of cell wall hemicellulose, which enhances fruit senescence. The study by Hatoum et al. (2016) established a relation between metabolic regulation during post-harvest storage and cellular respiration and stress.

In recent years, the Kiwifruit (Actinidia Lindl. spp.) has gained popularity in international markets due to its distinct appearance and the health benefiting nutrients such as vitamin C and fiber (Ward and Courtney, 2013). A total of 51 metabolites were detected during kiwifruit development and ripening (Nardozza et al., 2013). The content of soluble sugars and ascorbate significantly changes during ripening, which eventually determines the fruit quality and taste (Nardozza et al., 2013). Hence, the quality and flavor of Kiwifruit can be improved by targeting the metabolites that can render consumer acceptance. In Kiwifruit, application of synthetic cytokinin N-(2-chloro-4-pyridyl)-N′-phenylurea significantly increases fruit size, and affects the ripening processes by altering the accumulation pattern of metabolites such as amino acids, sugars, organic acids etc., (Ainalidou et al., 2015).

The quality and taste of orange (Citrus spp.) fruit depend on the composition of metabolites such as organic acids, sugars, vitamins, flavonoids, and carotenoids. The metabolomics study of orange bud mutant ‘Hong Anliu’ (accumulates higher levels of lycopene and sweeter than wild type) led to the identification of 130 metabolites that include acids, sugars, flavonoids, alkaloids, limonoids, coumarins, amino acids, and plant hormones (Liu et al., 2007; Pan et al., 2014). The flavor and the taste of ‘Hong Anliu’ sweet orange was determined by the higher levels of soluble sugars and lower levels of organic acids along with differential levels of flavonoids at ripe stage.

The infection of Candidatus Liberibacter asiaticus, causal agent of Citrus Huanglongbing disease, deteriorates juice quality (Slisz et al., 2012). The infection leads to severe decrease in glucose, fructose, sucrose and amino acids such as alanine, arginine, isoleucine, leucine, proline, threonine, and valine; whereas, it enhances the levels of citrate and phenylalanine. Heat treatment of fruit is widely used as a means to avoid fruit infection during post-harvest storage, which is well supported by metabolomics study. In a study, the heat treatment significantly decreased the content of organic acids and amino acid; however, it promoted the accumulation of metabolites such as 2-keto-D-gluconic acid, tetradecanoic acid, oleic acid, ornithine, succinic acid, myo-inositol, glucose, fructose, sucrose, and turanose, which reduces the risk of post-harvest infection (Yun et al., 2013). Recently, ABA is reported to serve as a regulator of citrus cuticular wax biosynthesis during fruit development (Wang et al., 2016).

In grape (Vitis vinifera), the fruit setting relies upon the abundance of metabolites, and is regulated by the reprograming of hormones and sugar metabolism pathways (Domingos et al., 2016). The effect of geographical distribution on grape metabolite content is well documented (Son et al., 2009). The grapes grown in the regions perceiving high sun light-low rainfall show enhanced content of sugars and amino acids, Na and Ca, along with low levels of organic acids, suggesting the role of extrinsic factors on grape fruit qualities. The metabolite abundance in grapes berry that is reported to be stage specific and cultivar dependent, regulates the ripening processes (Cuadros-Inostroza et al., 2016). Stilbenes are the major polyphenols present in the grapes that determine the quality of drinking wine. The MS analysis of grapes allowed the detection of several bioactive stilbenes like ampelopsin H, caraphenol, isohopeaphenol, trans-resveratrol, Z- and E-astringin, piceatanno, Z- and E-piceid, B pallidol and pallidol-3-O-glucoside and parthenocissin A (Flamini et al., 2015). The study focused on the polyphenolics content of the grape identified upto 450 compounds including anthocyanin, glycoside aroma precursors, flavanols and procyanidins, flavones and flavanones, phenolic acids and stilbenes. Particularly, this study allowed identification of several 100 compounds, which were used to build a new database of putative compounds (Grape Metabolomics).

Pear (Pyrus communis), a member of Rosaceae, is grown worldwide for its unique ‘melting’ texture. Japan is one of the largest producers of pears. The metabolomics analysis of pear fruit confirmed differential accumulation of ∼250 metabolites during fruit development and ripening (Oikawa et al., 2015). Ripening of pear fruit manifested accumulation of sugars (e.g., sucrose), sulphur-containing amino acids, phytohormone such as ABA and brassinosteroids. This study reported detection of 15 phytohormones including abscisic acid, auxin, brassinosteroids, gibberellins, jasmonic acid and salicylic acid. The blooming stage shows a substantial increase of the metabolites (amino acids and organic acids), which further decreases during fruit development.

Like pears, strawberry (Fragaria × ananassa) is rich in secondary metabolites such as flavonoids. The process of gain and loss of strawberry fruit flavors during evolution and domestication was illustrated by Aharoni et al. (2004). The cultivated species of strawberry predominantly contain terpenoids such as monoterpene linalool and the sesquiterpene nerolidol. Whereas, the wild species were rich in the olefinic monoterpenes and myrtenyl acetate. Surprisingly, these were absent in the cultivated species (Aharoni et al., 2004). The untargeted (GC-MS) and targeted (HPLC) based studies of strawberry fruits were employed at seven different stages of fruit development. The metabolic study revealed a shift in the metabolite content during fruit development and ripening. The strawberry ripening involved rise of free amino acid content, with change in sugar content, including substantial changes in other major metabolic pathways such as ester biosynthesis, shikimate, and tricarboxylic acid (Zhang et al., 2011).

The effect of biotic stress and the fungicide (to avoid biotic stress) on strawberry quality was evaluated by quantitative estimation of primary and secondary metabolites accumulated in the infected and non-infected fruits (Mikulic-Petkovsek et al., 2013). The Colletotrichum nymphaeae infection induces accumulation of sugars and reduces the organic acid content. The infected fruits displayed altered content of metabolites such as ellagic acid derivatives, flanonols, flavan-3-ols, oligomeric procyanidins and total phenolics. Recent work by Nagpala et al. (2016) revealed an increase in the polyphenol levels in white-fruited species of strawberry as a result of infection from fungal pathogens viz. Botrytis cinerea and Colletotrichum acutatum.

Forage and grain legumes contribute 27% of the world gross primary crop. The grain legumes alone cater 33% of required human dietary protein, contribute to food security and environmental sustainability (Graham and Vance, 2003; Ramalingam et al., 2015). Notwithstanding the extensively investigated model legumes, metabolomics studies in other legumes remain limited. Concerning model legume, investigation of the effect of rhizobial node factor (Nod) in Medicago revealed a decrease in oxylipins (Zhang et al., 2012). In another study, metabolic profiling of salt tolerant Lotus species uncovered a series of changes involving metabolic adjustments of shoot constituent for survival (Sanchez et al., 2011).

Stress conditions such as salinity and anoxia result in an accumulation of alanine, and its biosynthesis co-substrates such as glutamate and GABA, and succinate in soybean (Rocha et al., 2010b). Differential expression was also obtained for genes involved in nitrogen fixation and fermentation in root. Interestingly, a negative correlation was observed for the amino acid derived from glycolysis and the TCA cycle during water logging, and several TCA cycle enzymes were induced upon exposure to water logging (Rocha et al., 2010a). Likewise, an attempt to elucidate the metabolic changes associated with flooding stress in soybean led authors to identify a set of 81 mitochondria associated metabolites, thus suggesting a boost in concentrations of metabolites involved in respiration and glycolysis such as, amino acids, NAD and NADH coupled with the depletion of free ATP (Komatsu et al., 2011). Under drought and salinity conditions, metabolite phenotyping of four different Mediterranean accessions of lentil suggested a decrease in intermediates of the TCA cycle and glycolytic pathway (Muscolo et al., 2015). Importantly, this study yielded metabolite markers for specific stress; such as threonate, asparagine/ornithine and alanine/homoserine for NaCl, drought and salinity, respectively. Another study that aimed to assess the impact of water deficiency on Lupinus albus demonstrated plant stem serving as a storage organ for sugars and amino acids (Pinheiro et al., 2004). Importantly, tolerant plant accumulated significantly higher level of metabolites such as asparagine, proline, sucrose and glucose in the stem stelar region (Pinheiro et al., 2004). The authors proposed reorganization of nitrogen and carbon metabolism pathways in plants in order to tolerate salinity stress. In soybean, consistent increase in pinitol (sugar alcohol, osmoprotectant) was reported in the tolerant plant at both normal and drought-stressed conditions (Silvente et al., 2012). Similarly, accumulation of sucrose, free amino acids and soluble proteins was observed in tolerant soybean in response to water stress (Tripathi et al., 2015).

Cereals remain the prime source of nutrition worldwide owing to their grains rich in vitamins, minerals, carbohydrates and fats (Sarwar et al., 2013). Cereals have been widely studied in order to quantity variation in metabolites and their association with sequence variation (Chen et al., 2014, 2016). In rice, different research groups have harnessed the potential of metabolomics in order to explore the metabolites diversity between different varieties and natural variants (Kusano et al., 2007; Redestig et al., 2011; Gong et al., 2013; Hu et al., 2014, 2016; Kusano et al., 2015; Chen et al., 2016; Okazaki and Saito, 2016). Similarly, metabolomics studies in maize have allowed researchers to differentiate and subsequently select the superior genotypes with enhanced nutritional composition (Matsuda et al., 2012; Wen et al., 2014; Venkatesh et al., 2016). Recently, metabolomic approach has been utilized to survey chemical diversity between different maize and rice variety and its natural variants (Chen et al., 2016). In maize, drought stress is reported to be regulated by amino acid metabolism (Obata et al., 2015). Photorespiration is tightly regulated under drought as the two amino acids involved in this pathway, glycine and serine are rendered up-regulated. Further, accumulation of glycine and myo-inositol was reported to relate with grain size of maize under drought, implicating these metabolites as potential markers for identifying drought tolerant maize (Obata et al., 2015). Similar work in rice demonstrated drastic induction of certain compounds in tolerant plants such as allantoin, galactaric acid, glucose, gluconic acid, glucopyranoside and salicylic acid, which could be considered as metabolite markers to address drought stress in rice (Degenkolbe et al., 2013). As demonstrated in sorghum by Ogbaga et al. (2016), the plant’s ability to acquire and reorganize its metabolic status in order to cope with drought shows considerable variation within species. Under drought condition, sorghum variety having a greater tolerance to drought (Samsorg 17) accumulated sugars and sugar alcohols in comparison with less drought tolerant variety (Samsorg 40) that accumulated free amino acids. Marked abundance of soluble sugars with amino acids was also observed in the roots of tolerant barley plants under salinity stress (Shelden et al., 2016). Like drought, chilling stress is also known to induce accumulation of amino acids and carbohydrates. For example, chilling stress caused substantial changes in metabolic profiles of rice varieties viz. Nipponbare (Japonica) and 19-11 (Indica) (Zhang et al., 2016). The chilling tolerance of Nipponbare involved metabolic adjustment to activate antioxidation pathway by modulating key metabolites such as γ-glutamylisoleucine, γ-glutamylglutamine, 5-oxoproline, glycine, glutamate, adenine dinucleotide and putrescine (Zhang et al., 2016). Further, chilling stress activates glycolytic pathway, however, normal activity is resumed during recovery phase. In both wheat and barley, cold stress expedites the amino acid pool and induces the GABA-shunt genes to promote conversion of glutamate to GABA (Sutka and Snape, 1989; Mazzucotelli et al., 2006). It is well established that cereal grains accumulate flavones/flavone-glycosides, which protects plants from various stresses (Caasi-Lit et al., 2007). For example, rice produces plenty of flavone-glycosides to protect it from abiotic stress and herbivores (Adjei-Afriyie et al., 2000; Matsuda et al., 2012). However, examination of herbivore-induced defense system in maize showed an increase in azealic acid, N-hydroxycinnamoyl tyramines, phospholipids, tryptophan, and 1,3-benzoxazin-4-ones (Marti et al., 2013). Accumulation of resistance related metabolites is also reported during plant–pathogen interaction. For instance, a tolerant variety of wheat can accumulate a wide range of metabolites conferring tolerance such as coumaroylputrescine and coumaroylagmatine during fusarium head blight (Kage et al., 2016). Further, evaluation of these hydroxycinnamic acid amide compounds and their placement on metabolic pathways has led to the identification of an important gene agmatine coumaroyl transferase (ACT).

According to a report of the intergovernmental panel on climate change (IPCC), anthropogenic activity, deforestation and combustion of fossil fuel could boost CO₂ level upto 700 ppm by 2100 (IPCC, 2013)³. The CO₂ uptake and water availability are directly connected to photosynthesis and plant growth, and CO₂ sequestration by plants helps in maintaining terrestrial ecosystems (Weltzin et al., 2003; Reich et al., 2006; Arora and Boer, 2014). A recent study by Liu et al. (2016) has shown the impacts of elevated atmospheric CO₂ on plant growth rate, biomass and leaf area. Terrestrial plants and phytoplankton significantly utilize increased atmospheric CO₂ to increase their biomass (Lawlor and Mitchell, 1991; Schippers et al., 2004; Forkel et al., 2016). However, enhanced CO₂ levels might promote grass species in the long term (Smith et al., 2000), which is an encouraging finding concerning food crops such as cereal. Sustaining crop performance in the face of growing CO₂ levels remains a key challenge of 21st century agriculture. Therefore, studies are required to understand the metabolic composition and the relevant alterations on metabolome due to high CO₂ stress.

Lipids are an important constituent of cell membrane enclosing organelles and suborganelles, in which various biochemical reactions occur. Modern lipidomics has facilitated profiling of lipids to understand lipid dynamics and biosynthesis on exposure to a range of stresses (Kosma et al., 2010; Burgos et al., 2011; Szymanski et al., 2014; Hou et al., 2016; Li et al., 2016; Tenenboim et al., 2016).

Plants adjust their lipid structure according to varying environmental conditions (Tenenboim et al., 2016). For instance, cold tolerant plants increase the levels of desaturated glycerolipids to maintenance membrane fluidity (Sakamoto et al., 2004; De Palma et al., 2008; Degenkolbe et al., 2012). Freezing plants up to a sublethal temperature can induce the level of lysophospholipids, phosphatidic acid, and phosphatidylglycerol (Welti et al., 2002; Zhang et al., 2013a). In contrast, tolerance to heat stress involves an increase in saturated glycerolipids (Larkindale and Huang, 2004). Recently, MS based analysis revealed remodeling of lipids, antioxidants and galactolipids in tomato plant during high temperature stress (Spicher et al., 2016). A combined glycerolipidomic and transcriptomics study provided insight on lipid remodeling, and regulatory genes involved in lipid biosynthesis and heat stress management (Burgos et al., 2011; Szymanski et al., 2014; Higashi et al., 2015; Légeret et al., 2016; Narayanan et al., 2016a,b). Most strikingly, high temperature can induce dramatic increase of lipid antioxidant such as α-tocopherol and plastoquinone/-ol, and saturation of membrane lipids like galactolipids and phosphatidyl ethanolamine (Spicher et al., 2016).

Hypoxia represents another important type of abiotic stress that plant faces due to excessive watering or flood and leads to limited O₂ availability and increased salinity for a submerged plant (Rajapakse et al., 2009). According to Xie et al. (2015), plant cell synthesizes and accumulates unsaturated ceramides, a class of sphingolipids under hypoxia. More recently, root lipidic content of two barley genotypes was examined to understand the mechanism underlying their tolerance to salinity stress (Natera et al., 2016). Phosphorus availability during stress is known to directly affect the membrane lipid texture. For instance, Arabidopsis grown in phosphorus-deficient condition can induce replacement of phospholipids with non-phosphorus SQDG class lipids (galactolipids) (Essigmann et al., 1998; Härtel et al., 2000). These galactolipids are mostly associated with plastic thylakoid membrane, and phosphorus deprivation causes its enrichment in the roots extraplastidic membrane for survival.

Compared to abiotic stress, lipidomics studies of biotic stress are scanty. During biotic stress, lipid peroxidation occurs due to formation of reactive oxygen species (ROS), ultimately leading to program cell death (PCD) (Zoeller et al., 2012). A recent report suggests inositol phosphorylceramide synthase as instrumental to coordinate the PCD, a mechanism acquired by plants for its self-defense to limit biotrophic pathogens (Wang et al., 2008). Lipid peroxidation results in formation of jasmonate and oxylipins, which are signaling molecules during plant immune response (Shah, 2005). Recently, nearly 100 membrane-associated lipids were quantified in response to methyljasmonate and cerium, suggesting increase of lysophosphatidylcholine, phosphatidic acid and phosphatidylcholine associated with PCD (Yang et al., 2008). Plant–pathogen interaction also impacts upon plant cuticle that serves as a first physical barrier limiting pathogen invasion along with protecting plants from other physical damages (Jenks et al., 1994). The cuticle layered over epidermal cells, is mainly composed of wax and cutin (Samuels et al., 2008; Mazurek et al., 2017). The cutin is mainly composed of hydroxylated C₁₆ and C₁₈ (Cheng and Walden, 2005; Kunst and Samuels, 2009). The permeability of cuticle relies upon its composition, which can restrict the invasion of fungal pathogens such as Botrytis cinerea (Saladié et al., 2007; Curvers et al., 2010). The role of cuticle in relation to plant defense against pathogen invasion is well described in the recent articles (Chassot and Métraux, 2005; Bessire et al., 2007; Chassot et al., 2008; Raffaele et al., 2009; Reina-Pinto and Yephremov, 2009; L’Haridon et al., 2011; Buxdorf et al., 2014; Serrano et al., 2014; Lara et al., 2015; Fernández et al., 2016).

Metabolites are the ultimate downstream factors that regulate and decide the cell fate; hence the content of metabolite directly affects organ physiology and often signifies the quality of fruits. The improvement of gene annotation is important to validate the gene function. In fact, it facilitates the use of these genes in the field of crop sciences to improve the quality and yield. This section describes significant leads achieved in the field of crop species through metabolic engineering.

The last decade has witnessed tremendous advancements in technologies followed by their deployment in understanding different facets of biology to understand the complex biology of desired traits. Knowledge-based plant breeding (KPB) utilizes all the meaningful information inferred from analysis of the large-scale data pertaining to genome, epigenome, transcriptome, metabolome, and proteome that collectively lead to a particular phenotype. The data generated from these ‘Omics’ approaches enable better understanding of the systems biology of the trait, so far mostly contributed by systems genetics and genomics, and to some extent transcriptomics. The other ‘Omics’ approaches, including metabolomics and proteomics have also started contributing toward generating important information, which can become an integral component of KPB, thereby strengthening the approach further for achieving higher genetic gain.

Technological advancement has contributed to improve the efficiency of plant breeding techniques via precise selection of desired plants. An easy access to the different ‘Omics’ platforms will cause a paradigm shift in breeding process by facilitating plant selections based on the genome-scale information generated at different levels of biological processes. The plant breeders will gradually embrace these developments, which in turn will help them to make informed decisions.

Advances in plant metabolomics in recent time has allowed the precise selection of desirable traits along with offering opportunities to undertake metabolic engineered plants. The shift of technology from single metabolite analysis to high throughput assays generating footprints of a variety of metabolites in one go has paved the way for discovery/construction of better models for metabolite networks, and the identification of robust biomarkers. In the last decade, the implementation of metabolomics in conjunction with other omics technologies has not only uncovered a plethora of known as well as novel metabolites, but also allowed to determine their specific contribution toward improving key plant attributes such as quality, yield, shelf life, etc. To this end, high throughput genotyping/sequencing platforms based on NGS technology has been a tremendous support as a cost-effective and high-throughput means to elucidate the architecture of metabolic traits. The newly created avenues such as GWAS, GS and EWAS that allow efficient integration of metabolite profiling could provide a great impetus to metabolomics assisted breeding. We anticipate that the integration of metabolomics and the other omics tools greatly improves the ability of a plant breeder in order to design and develop agronomically superior plants, thus enabling rapid development of high-performing crop genotypes that adequately meet the challenges of 21st century agriculture.

RK, AB, MKP, and AK designed the article. RK and AK wrote the article. AB, AP, RK, and MKP corrected the article and finally all authors read and approved the final article.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.