Genomics is the study of recombinant DNA, DNA sequencing techniques, bioinformatics, examining the structure, function, evolution, mapping, epigenomic, mutagenomics of genes and genomes. ⁸ DNA sequences and annotations, among other genomic tools, are now deposited in publicly available databases. ¹⁷ The progressive improvement of computers and networks is required for the storing and handling of these resources. In natural and engineered strains, genomics enables high-throughput DNA sequencing and large-scale bimolecular modelling of metabolic and signalling networks. ¹⁸ In this era, genomic technologies have been applied to improvement of crop breeding, plant promotion and protection. A body of evidence has shown that bacteria and fungi co-exist with their host plants, thus promoting plant growth and protection. ^{19 , 20} Tremendous progress has been made in structural genomics indicating morphological and genetic maps to find features of interest.

Countless studies have expressed several genes related to stress tolerance, ²¹ nitrogen fixation, ²² for mineral acquisition (Fe, P, etc.), ²³ phytohormone production (IAA, GA, etc.), adhesion, and other colonization related genes, ²⁴ which are significantly important in microbial endophytes existence. The introduction of beneficial microorganisms to crops might lead to plant growth along with antagonistic activity against plant diseases. ²⁵ It is worth noting, that the agriculture industries are becoming increasingly involved in finding more sustainable, greener, and environmentally friendly products, which aligns to the 17 Sustainable Development Goals (SDGs). In addition, this is also in response to the worldwide problem of pesticide exploitation and overuse. ²⁶ Microbial products can be employed as biofertilizers, plant strengtheners, phyto-stimulators, and biopesticides, depending on their mode of action and effects. ²⁷ Tremendous progress has been made in improving the scale-up and bioprocess development of microbial inoculants. The application of genomic technologies, particularly microbial inoculants, have massive benefits. ¹⁹

Sequencing technologies has cemented a “gold standard” for the detection of both known and unknown variants in genomics. Depending on the intended downstream application, different strategies are utilized for each phase. Subsequently, nucleic acid-based sequences will go through cleaning, filtering, assembly, alignment, variant calling, annotation, and functional predictions. ²⁷ Shotgun sequencing was one of the most popular techniques utilized with long strands of DNA as well as complete genomes. The conventional chain-termination approach, also known as the “Sanger method,” which is based on the selective incorporation of chain-terminating dideoxynucleosides by DNA polymerase during in vitro DNA replication, was the technology behind shotgun sequencing for a significant portion of its history. In addition, Sanger sequencing is time consuming hence the use of high throughput technologies nowadays, e.g., next generation sequencing (NGS). High-throughput sequencing or microarray hybridization, as well as bioinformatics, are essential genomic analysis methods. ²⁸

Despite the massive benefits, microbial inoculants are based on various modes of action, and their interactions are influenced by environmental factors. Another key concern is that microbial inoculants are intended for a wide range of crops globally, they thus might react differently in different species. ¹⁹ Moreover, to obtain an “ideal” inoculant, culturable microbial species require optimal growth conditions and in-depth understanding of plant/host relationships. ²⁹ To address this, fermentation and formulation procedures need to be improved further. Government legislation remains an obstacle in safety aspects of commercialized microbial inoculants. ³⁰ Moreover, microbial strains may also act differently in culture than they do in their native habitats. As a result, it is critical to create cultivation-independent methodologies for studying microbial communities. ³¹ Genomic technologies are taking a leap in the right direction by developing innovative biotechnology strategies in plant growth and protection. ¹⁹ Unfortunately, genomics technologies can only say whether genes are present or not. To address this, transcriptomics can determine when and where each gene is turned on or off. The RNA sequence mirrors the DNA sequence from which it was transcribed, ultimately gene expression is the logical follow-up in research advancements. To gain insight into plant breeding, biotechnological and genetic engineering, a multi-disciplinary approach linking structural, functional, mutational and comparative genomics is required ( Figure 2).

Compounds have an impact on gene expression, which might provide insight into their functional and toxicological features. Expression investigations on disease and disease-free clinical samples may lead to the discovery of new biomarkers. Importantly, gene expression analysis can also be utilized to learn more about the physiological effects of plant genetic manipulation. ³⁹ Outdated techniques such as Northern blotting, serial analysis of gene expression have massive drawbacks that make them inappropriate for analysing significant numbers of expression products at the same time. The development of DNA microarray technology has recently resulted in a significant increase in the sensitivity and throughput of expression screening. ⁴⁰ Despite a number of technological and methodological challenges, microarray research has resulted in major improvements in our understanding of biological processes. It can be a low-cost method of identifying the most likely interesting subsets of samples that will produce results in other technological tools. ¹⁶ High-throughput technologies have overtaken the omics world and transcriptomics is no exception with the introduction of whole-genome and next-generation transcriptome. Hence, transcriptome profiling is easily accomplished as a result of use of deep-sequencing technology.

Microarrays have now been surpassed as the preferred approach for gene expression profiling by RNA-seq. Furthermore, it also allows for much more exact determination of transcript amounts and isoforms than other approaches. In a recent study conducted by Ahmad and co-workers, ⁴¹ based on RNA sequencing, the molecular basis of Brassica rapa plant resistance to Hyaloperonospora brassicae (a fungus-like pathogen) was determined. The elevation of PR genes resulted in a physiological response. The loci Bra003774 and Bra025730 in the PR gene family showed a two-fold increase in expression, indicating that WRKY22 and PR1 are overexpressed. PR1 was identified as a delayed response factor to the assaulting pathogen due to its end location in defence responses. PR1, zinc finger protein metabolism was the second most strongly active activity in Brassica cells, with 42% of genes showing increased expression. The plant's transcriptome profile was identified in order to better understand how the host reacts to the pathogenic onslaught and ensure its survival. The authors suggested that the study will be extremely beneficial to global plant protection efforts. ⁴¹

In another study, heavy metal particularly cadmium (Cd) tolerance was determined by using transcriptomic profiling in the Calotropis gigantea plant. To deal with Cd, the leaves triggered a number of Cd detoxification pathways, including overexpression of genes involved in Cd transport such as absorption, efflux and enzyme activation. The phytoremediation capacity of C. gigantea in Cd-contaminated soils was investigated using Cd tolerance parameters and molecular mechanisms. ⁴² RNA-sequencing was used to exam the reactions of two varieties inoculated with Alternaria alternata, a resistant type and a susceptible type. A total of 2235 differentially expressed genes (DEGs) in disease-resistant cultivars and 4958 in susceptible cultivars were examined. Furthermore, the expression of DEGs as a defensive response, detection of resistance genes and signal transduction pathways, which will aid in the exploration of resistance mechanisms in response to A. alternata infection, were determined. ⁴³ Transcriptomic research has aided in the identification of key functional genes and pathways that may show memory behaviour. Moreover, studies have also been useful in elucidating the function of memory processes in the modulation of ABA-related gene expression in response to drought. ⁴⁴

RNA-Seq, the NGS based technology, has become the main choice to measure gene expression levels and comparing gene expression patterns at unprecedented resolution. Transcriptomics identifies the genes expressed in response to biotic and abiotic stress. A basic reference-based RNA-Seq data analysis pipeline includes pre-processing (removing low-quality sequences to improve alignment), alignment of raw reads with the reference genome, transcript assembly, and detection of differentially expressed genes. ⁴⁵ The new tuxedo suite has been developed because it requires less memory amongst other advantages. It includes three tools for transcript-level analysis of RNA-Seq data, including HISAT2 for spliced alignment of reads to the genome, assembly of transcripts based on mapping to the genome, including novel transcripts, and quantification of these transcripts. ⁴⁶ The final step for the new tuxedo suite is to identify differentially expressed genes and transcripts using Ballgown. ⁴⁷ Pathway approaches analyze databases and gene expression data to identify the significantly impacted pathways in a given condition. ⁴⁸

Despite the potential benefits, there has been records of noticeable downfalls of using transcriptomics. Transcriptomics is not suitable for recognizing adaptive stress response genes. This is due to the massive change in gene expression which may not always translate into a large influence on fitness, enormous gene effects are uncommon, protein activity is most relevant to fitness, and mRNA abundance is an inconsistent predictor of protein activity. It can be computationally difficult to annotate sequences accurately and interpret data, especially when there is no reference genome available. ⁴⁹ Furthermore, biases established during the creation of the cDNA library and sequence alignment can have an impact on the quantification of transcripts. Reproducibility can be hampered by differences in read depth between sequencing platforms. Although RNA-seq has become more economical, many laboratories still cannot afford it. A key component in transcriptomics is that the start-up expenses are high, as well as the cost of each sequencing reaction, which depends on the read depth. ⁵⁰ However, the benefits outweigh the shortfalls, especially when advanced multi-omics is applied. Our better understanding provides new knowledge regarding the relationship between gene expression and fitness in demanding environmental settings.

Plant proteomics research has advanced and brought improvements in protein extraction and separation to larger scales, moreover becoming a complement technique to transcriptomics and metabolomics. ⁵⁶ Though proteomics could be a powerful approach it requires a series of steps from protein separation and identification to protein functional analysis and bioinformatics inferences involved in crop stress response. ⁵⁷ As opposed to the old conventional proteomics techniques that were mostly chromatography based, recent techniques including SDS-PAGE, 2-DE, and two-dimensional differential gel electrophoresis (2D-DIGE) were developed and are gel based. ³³ On the same level, protein microarrays/chips utilize minimal sample for protein expression analysis. Additionally, advanced isotope labelling has played a very important role in quantitative proteomic analysis especially in crop improvement studies of abiotic stress adaptation. ³³ For key molecular weights identification MALDI-TOF, electrospray ionization (ESI), and collision-induced dissociation (CID) are utilized. ⁵⁸

Several bioinformatics tools and databases have been developed for various applications including genomics analysis, transcriptomics, proteomics and metabolomics, while a comprehensive list of each approach has been provided separately. ⁵⁹ Proteomics studies generates a large data set using various analytical techniques that are spectrometry or gel based, however such analyses are mainly dependent on several bioinformatic software tools that process, interpret, mine, assess and annotate functions. A bioinformatic tool used for data interpretation obtained from 2-DE gels include the SWISS-2DPAGE. Moreover, multiple databases exist that provide evolutionary and functional annotations of orthologs, predict protein’s subcellular locations such include OrthoDB, OMA orthology, UnitprotKB, WoLF PSORT. ⁶⁰ Various studies have been carried out on the identification of subcellular locations of various proteins to gain guidance for further experimental investigations. ^{61 , 62} Gene Ontology (GO) term identification provides terminology for biological processes and group associated genes according to various functional terms that describe the “biological process”, “molecular function” or “cellular component”. ⁶³ For larger data sets and systematic approaches, tools such as MaxQuant, Proteome Discoverer ⁶⁴ and X!tandem are used for proteomic annotations. ⁶⁵ Furthermore, several bioinformatics databases exist that provide proteomics analysis as indicated in Table 1.

Functional proteomics has identified different biochemical components such as reactive oxygen species (ROS) including such as quinone reductase, g-glutamylcysteine synthetase, dehydrins, and dehydroascorbate reductase reported in tomato and sunflower crops. ⁷³ Important chaperones, such as heat shock proteins, have been identified during proteome functional analyses in wheat and sugarcane. ⁷⁴ Quantitative proteomic studies have also been shown to play a crucial role in various crop responses against abiotic stresses. For example, the use of the iTRAQ method in differential expression of proteins in potato and in two coconut varieties under abiotic stress. ^{75 , 76} Furthermore, iTRAQ combined with the LC-MS/MS technique has proven to be an effective method widely used for fast identification and quantification of complex protein mixtures in abiotic and biotic stress responses in plants. ⁷⁷ A combination of numerous proteomics approaches including MALDI-TOF, SDS-PAGE, MS, 2-DE, and PMF have been utilized to study response to diverse abiotic stresses in various crops such as rapeseed, soybean, wheat, sugarcane, and cotton. ^{74 , 78}

There are various limitations that exist in proteomics analysis that normally originate from the initial stage that includes preparation and extraction of samples especially in recalcitrant crops with high content of polyphenols, ⁷⁹ lack of a detailed number of proteins observed and inadequate tools in proteomic data analysis. Furthermore, challenges in peptide separation is a major inhibiting factor to downstream analysis, mass spectrometry-based analysis and bioinformatics tools. ⁸⁰ Another limitation includes lack of a one-step quantification approach that can fully integrate a multi-functional system allowing comprehensive quantification of a wide spectrum of proteins from model and non-model plants. ⁸¹

Several analytical techniques are used to generate metabolic profiles of a given plant sample, including thin layer chromatography (TLC), gas/liquid chromatography-mass spectrometry (GC/LC-MS), liquid chromatography-electrochemistry-mass spectrometry (LCEC-MS), nuclear magnetic resonance (NMR), direct infusion mass spectrometry (DIMS), Fourier-transfer infrared (FT-IR), and capillary electrophoresis. ^{82 , 85} These methods depend on the approach's sensitivity, selectivity, speed, and accuracy. The methods CE-MS, GC-MS, LC-MS, and NMR are typically employed to profile metabolites from various plant materials. ⁸⁶

Nuclear magnetic resonance is considered to be fast and selective with respect to metabolite profiling as opposed to mass spectrometry-based approaches. ⁸⁴ However, a combination of NMR (semi-quantitative) and MS (quantitative) is advisable since it offers a better understanding of metabolome quantification due to the complementarity of these two approaches. ¹⁸ Metabolomics, combined with other omics approaches such as genomics, transcriptomics, and proteomics serve as a powerful approach that offer solid insights on molecular response pathways, plant biochemical and physiological mechanisms; and their related functions, ⁸⁷ thus ultimately result in its relevance in food security.

Several plant metabolomic bioinformatic tools are used to analyse metabolic raw data generated by various analytical tools for data processing, mining, annotation, interpretation and statistical analysis. ⁸⁸ Different web-based programs are used for data pre-processing of metabolites those includes XCMS, METLIN, AMDIS, MeltDB, MetaboAnalys, MetAlign, MZmine 2, and AnalyzerPro for varying analytical techniques. XCMS is a web-based program that assists the user with data processing and statistical analysis, ⁸⁹ while METLIN is mainly used for metabolite annotation. ⁹⁰ MeltDB tool is used for data assessment, processing, and statistical analysis. ⁹¹ Lastly, AMDIS, KNApSAcK and KOMICS are mainly used for data processing and metabolites present in crop species. ⁹²

Plants produce metabolites (secondary) as defence mechanisms against abiotic and pathogen stress. Several studies have reported identification of various metabolites in cereal crops including rice, maize and barley in response to various biotic stressors. ⁹³ Rice cultivars demonstrated identification of several metabolites using GC-MS against a pathogen gall midge biotype 1 (GMB1). ⁹⁴ Several metabolite analyses were reported in wheat, maize, tomato, and soybean crops in response to drought, cold and heat stress. ^{95 , 96} Therefore, metabolomics offers a significant advantage compared to its counter-omics approaches such as transcriptomics and proteomics, in that it reveals the downstream products synthesized by genes and expressed by proteins. ⁸³ However, to comprehensively classify metabolites according to their various chemical/physical properties relevant preparation methods and instrumentation need to be utilized for enhancing metabolite performance. ⁹⁷

There are several disadvantages that impede the full characterization of metabolites from a given organism, including the complex physicochemical properties that cannot be characterized by single or limited analytical techniques, hence combined multiple analysis tools are recommended for a full metabolome analysis. ⁸² Another limitation in creating metabolomics profiles is the consistency of growth conditions and preparation of plant samples due to the instability of metabolites undergoing multiple modifications such as hydrolysis or oxidation. Improper harvesting, handling, extract preparation, and sample storage are the most likely sources of instability in plant metabolite analysis. ⁹⁸ In addition, simple metabolomics analysis requires the use of expensive equipment compared to genetic analysis, hence there are still limited data on these studies. Metabolomic analysis in plant pathology is challenging as it cannot distinguish between plant and microbial metabolites resulting from the interaction between (host and pathogen) plants and microbes as a metabolic response pathway. ⁹⁹ Limited universal metabolite-specific standards and reference compounds to assist in identifying metabolomes. ¹⁰⁰

The introduction of high-throughput sequencing (HTS) has revolutionized the field of microbial ecology. ¹⁰⁵ Previously, identification of bacteria and fungi was largely done using culture dependent techniques, which are inefficient as some microorganisms cannot be cultured on artificial media. Traditional methods for detecting viruses, such as polymerase chain reaction (PCR) and enzyme-linked immunosorbent assay (ELISA) (PCR), typically target a few viral species and require prior virus knowledge to detect. ¹⁰⁶ HTS sequencing allows for the study of microbes by sequencing the genetic material directly on the environment, ¹⁰⁷ avoiding the need for culturing, termed metagenomics. ¹⁰⁸

The most exhausted high-throughput sequencing application in microbial ecology is target-gene amplicon sequencing. The 16S rRNA gene is found in all bacteria and has either high or low sequence variability. ¹⁰⁹ This gene contains nine hypervariable regions (V1–V9) that show significant sequence diversity among bacteria. ¹¹⁰ Although no single hypervariable region can distinguish between all bacteria, regions V2, V3, and V6 have the greatest discriminating power and contain the most information in plants. However, Kembel et al. ¹¹¹ argued that the 16S rRNA marker may be inconvenient due to variable copy numbers, with some taxa having up to 15 copies of this gene. As a result, there may be taxa overrepresentation in the final set of sequences, resulting in biased community structure estimations. Nevertheless, the benefits of using 16S rRNA gene markers outweigh the inconvenience, which is why amplicon sequencing is widely used in surveys.

The internal transcribed spacer (ITS) of fungi is a non-coding region that is highly polymorphic and contains the taxa that can separate sequences into species. ¹¹² It is found in the ribosomal RNA operon and ranges in length from 450 to 750 bp. ¹¹³ The benefit of using ITS for-DNA barcoding is that it has been used in numerous studies and has updated reference sequences in the NCBI database. ¹¹⁴ ITS1, ITS2, ITS3, and ITS4 are the most used oligonucleotides for sequence-based fungal classification at the species level. ¹¹⁵ Therefore, amplicon sequencing forms the basis in characterizing bacteria and fungi and is an ideal tool to begin studying the microbiome as it is reasonably inexpensive.

Shotgun metagenomics has transformed microbiome detection and discovery by enabling untargeted characterisation of entire microbiomes. It has revolutionised plant science by enabling the untargeted, simultaneous detection of multiple viruses, bacteria, archaea and fungi regardless of their genomic nature and can deliver species level identification. ¹¹⁶ This includes annotations for metabolic and functional processes. Shotgun-based approaches typically generate large, sequenced data. ¹¹⁷ However, their ability to accurately group DNA sequences or assembled sequence contigs into single-species genomes has remained limited. ¹¹⁸ Recently a powerful approach has been introduced for metagenomic analysis that can address the challenges with shotgun. ¹¹⁹ Hi-C is a technology that provides direct and quantitative measurements of DNA sequences from shotgun sequencing. ¹²⁰ This technology yields higher numbers of high-quality genomes and captures strain-resolution insights. Hi-C eliminates the need for guesswork in metagenomic studies as direct evidence sequences. ¹²¹ The technique allows for accurate attribution of plasmids, phages, antibiotic resistance genes and other mobile genetic elements to host cells. ¹²⁰

Bioinformatic analysis of plant microbiomes demands the use of pipelines that effectively analyze the high throughput data generated by shotgun and target sequencing for microbial profiling and gene annotations. Several pipelines for analysing 16S rRNA gene sequencing data are available, including Quantitative Insights into Microbial Ecology (QIIME2), ¹²² MOTHUR, and Metagenomics - Rapid Annotation using Subsystems Technology (MG-RAST). ¹²³ CLC Genomics workbench ¹²⁴ and MEGAN. ¹²⁵ The comparison between the diversity and taxonomic compositions generated by MG-RAST and QIIME using samples was investigated. ¹²⁶ The study concluded that QIIME2 produced compositional assignments that are more accurate. Similarly, Plummer ¹²⁷ found that QIIME2 performance outweighed other platforms. Analysis with the platforms cluster similar sequences into operational taxonomic units (OTUs) or amplicon sequence variants (ASVs) and assign them to a taxonomic reference database, such as GTDB, the most updated bacterial database. ¹²⁸

Essentially the data analysis of metagenome shotgun involves assembly, which includes joining reads into contig sequences that are larger. Newbler, AMOS, MIRA methods were used in assembling metagenomes at an early stage of metagenomics. ¹²⁹ However, due to the low coverage and complexity of metagenomic reads, de novo assemblers, beginning with SOAPdenovo, were preferred for reference mapping. Genovo, ¹³⁰ Meta-IDBA, ¹³¹ MetaVelvet, ¹³² MAP, ¹³³ and Ray Meta are among the improved assemblers now available. ¹³⁴ Following assembly, microbial communities are binned and taxonomically characterized. Functional annotations mainly use BLASTX search against several databases, including SEED, KEGG, EggNOG and COG. ⁵⁷ Further graphical and statistical analysis is carried out using R software in conjunction with RStudio software and various microbiome-related R-packages. ¹³⁵

Metagenomics has been widely applied on different agriculturally important plants and on different microenvironments including the root and leaves. This has also been conducted in response to both abiotic and biotic stresses. For instance, Xu and colleagues ¹³⁶ revealed the role of iron metabolism in drought-induced rhizosphere microbiome dynamics using metagenomic tools. Similarly, patterns in the microbial community and the functional genes associated with salt-tolerant plants and root associated microbiomes of potato grown in Alfisols have been conducted in response to abiotic stress. ^{137 – 139} Metagenomics studies in response to different bacterial, fungal, and viral diseases have also been conducted in pear fruits, citrus, grapes, tomato, and bananas. ^{138 – 142}

Plant chloroplast and mitochondrial genomes, as well as bacterial 16S rRNA sequences, are similar, making bacterial rRNA gene amplification and sequencing challenging. As a result, the majority of the sequence data is from plant sequences, which are irrelevant to the study. Zarraonaindia et al. ¹⁴¹ showed that chloroplast sequences were the most problematic for Vitis vinefera, accounting for up to 98 percent of the total, as opposed to the mitochondrial sequences. It has been proposed that universal peptide nucleic acid (PNA) clamps can obstruct the host plant derived mitochondrial (mPNA) and plastid (pPNA) sequences at the V4 16S rRNA locus. ¹⁴³ However, universal PNA efficacy was tested on Arabidopsis thaliana and the sorghum plant and was discovered to be effective in preventing the amplification of non-targeted sequences. ^{38 , 144} Additionally, another primer set has also shown to reduce non-targeted plant DNA amplification when performing metagenomic research on microbial endophyte communities ^{. 10 The current study has highlighted some shortfalls of using a single omics approach and Table 2 provides a multidisciplinary approach by supplementing some limitations with other omics approaches, hence providing a multi-layered information system.}