The emergence of “innovative biology” is accompanied by the birth/innovation of other sciences, such as computational biology and bioinformatics, which have a combined interface of molecular biology. Due to the large datasets generated, its management and storage become critically important. Therefore, different databases came into existence, which organise a large amount of biological information stored and processed to permit the scientific community access (Ritchie et al., 2015). The increasing amount of data has been abetted by an increase in the number of biological databases (Pevsner, 2015). Usually public databases accumulate big amounts of information, and they are categorised into primary and secondary databases. The primary databases are composed of the findings of experimental data that are reported without any critical analysis related to previous publications (Luscombe et al., 2001; Prosdocimi, 2010). However, in the secondary databases, there is a collection and explication of data, called process of content curation. Besides various functional databases such as the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Reactome that allow analysis and explanation of metabolic maps. Various primary databases like DNA Database of Japan (DDBJ), GenBank at the National Center for Biotechnology Information (NCBI), and European Molecular Biology Laboratory (EMBL) remained as the main databases of nucleotide sequences and proteins. International Nucleotide Sequence Database Collaboration (INSDC) being the parent organisation of these databases and sharing among each other the deposited information daily (Prosdocimi et al., 2002; Amaral et al., 2007; Pevsner, 2015).

Last 2 decades have witnessed great advancements in molecular biology, data analysis procedures were established at a fast pace to enable the interpretation of the large amount of information produced mainly by DNA sequencing technologies that produced the exponential amelioration of genomics, transcriptomics and proteomics information. Biological data of genomics/proteomics although considered to be the recent domains, have emerged interdependently and created a historical impact on the available information coupled with innovations in computational resources, resulted in huge biological data and data analysis that can enhance and intensify the developments in medical science (Verli, 2014). In the current modern times ‘-omics’ suffix include the genomics, transcriptomics, proteomics, phylogenomics, metabolomics and metagenomics, associated with large-scale biological data and the allied bioinformatics analysis. The emergence of newest high-throughput sequencing innovations, starting with improvements in Sanger sequencing, innovations in NGS technologies and next-generation proteomics, resulted in emergence of novel findings in the clinical settings (Zhou et al., 2010).

DNA sequencing plays a crucial role in the progression of molecular biology, not only changing the genetic landscape of genome designs but also opening up new opportunities in therapeutic arena and personalised medicine

Generally, Genomics is the domain that aims to uncover and explore structure, function, and innovative realm of genomes applying bioinformatics tools to explore sequenced genomes. (Altmann et al., 2012).

Paul Berg’s (Jackson et al., 1972), Frederick Sanger’s (Sanger and Coulson, 1975), and Walter Gilbert’s (Maxam and Gilbert, 1977) pioneering work on DNA sequencing enabled several developments, including the advances that opened up completely new potentials for DNA analysis, Sanger’s ‘chain-termination’ sequencing technology, more commonly known as Sanger sequencing (Sanger et al., 1977). Further technological advancements steered in the rise of DNA sequencing, led to the development of the first automated DNA sequencer (ABI PRISM AB370A) to be released in 1986, allowing drafting of the human genome to be completed during the next decade (Venter et al., 2001). These new methods are meant to supplement and eventually replace Sanger sequencing Figure 1. This technology is commonly known as next-generation sequencing (NGS) or massively parallel sequencing (MPS), which encompasses a wide range of methodologies. It is feasible to create huge amounts of data & each instrument runs in a faster and more cost-effective manner using this technology. The Next Generation Sequencing market is currently developing and expanding, with the world-wide market expected to reach 21.62 billion US dollars by 2025, up around 20% from 2017 (BCC Research, 2019). As a result, multiple brands are currently competing in this business, including BGI Genomics, Illumina, Ion Torrent (Thermo Fisher Scientific), PacBio and Oxford Nanopore Technologies etc. All of them provide distinct approaches to the same query: the generation of sequencing data. Second-generation sequencing relies on large parallel and clonal amplification of molecules (PCR, polymerase chain reaction) (Shendure and Ji, 2008), whereas third-generation sequencing depends on sequencing of single-molecules without a preceding clonal amplification (Schadt et al., 2010; van Dijk et al., 2018; Ameur et al., 2019). Although the process of NGS include various steps:

1) NGS library Preparation: A library comprises DNA/RNA fragments that denotes the full genome/transcriptome or a region of interest in next-generation sequencing. Each NGS platform has its own unique features, in general, the production of an NGS library begins with fragmentation of the DNA/RNA, followed by the connection of sequence adaptors to fragments to permit enrichment of those fragments. The sensitivity and specificity of a good library should be high. This implies that all relevant fragments should be properly represented in the library and that there should be no random errors (non-specific products). It is easier said than done, though, because genomic areas are not all equally susceptible to sequencing, making the creation of a sensitive and specialised library difficult and cumbersome (Aird et al., 2011).

2) NGS Platforms

Platforms for Second-Generation Sequencing

The category of cyclic-array sequencing technologies (Amaral et al., 2007) includes second-generation systems. The production and library amplification (made from RNA/DNA samples), clonal growth, sequencing, and investigation are all part of the core workflow for second-generation platforms. Ion Torrent and Illumina are the two most well-known sequencing firms for second-generation sequencing systems (Kircher et al., 2011; Quail et al., 2012).

3) Platforms for Third-Generation Sequencing:

The ability to avoid limitations of PCR-based methods, such as nucleotide misincorporation by a polymerase, formation of chimaera and drop-outs of alleles resulting in an false homozygosity call, was made possible by 3^rd-generation NGS technology (Thompson and Steinmann, 2010). The Helicos Genetic Analysis System was the first commercial third-generation sequencer (Pushkarev et al., 2009). The Pacific Biosystems (PacBio RS II sequencer) established the notion of single-molecule real-time (SMRT) sequencing in 2011 (McCarthy, 2010). Furthermore, this method allows for the sequencing of lengthy reads (up to 30 kb on average). Individual DNA polymerases are coupled to zero-mode waveguide (ZMW) wells, which are nanoholes where a single DNA polymerase enzyme molecule can be put directly (McCarthy, 2010). PacBio has released the Sequel II System, which claims to cut project costs and timelines by up to 175 kb with highly accurate individual long reads (HiFi reads) compared to previous versions (Pereira et al., 2020).

Merker and co-workers demonstrated initially to use a PacBio System for sequencing of long-read genomes to find a pathogenic variant in Mendelian disease patients, indicating that this method has a lot of potential for identifying structural variation (Merker et al., 2018). The Chromium instrument, which uses gel beads in emulsion (GEMs) technology, was released by 10X Genomics in 2016 (Pereira et al., 2020). The benefit of GEMs technology is that it cuts down on time, beginning material, and prices (Zheng et al., 2016; Zheng et al., 2017; Pereira et al., 2020). With low false positives and high throughput, the chromium system can also perform single-cell genomic and transcriptional profiling, immunological profiling, and chromatin accessibility studies at single-cell resolution. As a result, intriguing new applications are emerging, particularly in the areas of epigenetics research, de novo genome assembly, and long sequencing reads (Delaneau et al., 2019; Laurentino et al., 2019; Wang et al., 2019).

4) Innovative Bioinformatics approach: Sequencing platforms are improving, and it is now possible to sequence the human genome in as little as a week or two. Thus, the huge data generated necessitates bioinformatics and computational expertise to organise, analyse, and infer NGS data. As a result, NGS bioinformatics is undergoing significant development, which can only be aided by improving computational capabilities (hardware) as well as algorithms and applications (software) to streamline all required steps: from processing of raw data to detailed data analysis and variant interpretation in a clinical setting.

Analysis of the NGS data: NGS bioinformatics is usually classified into three categories: primary, secondary, and tertiary analysis (Pereira et al., 2020).

The primary data analysis includes the identification and evaluation of raw data (signal analysis), the target of the generation of legible sequencing reads (base calling), and the estimation of base quality (Ledergerber and Dessimoz, 2011). This main analysis often produces a FASTQ file (Illumina) or an unmapped binary alignment map (uBAM) file (Ion Torrent).

Secondary analysis, which involves read alignment against the reference human genome (usually hg19 or hg38) and variant calling, is the next step in the NGS data analysis workflow.

Read alignment, which includes aligning sequenced fragments (processed data) against a reference genome, or de-novo assembly, which involves constructing a genome from basic without the use of external data, are two options for mapping sequencing reads. The availability or absence of a reference genome could be enough to decide between one technique and another. Nonetheless, reference sequence mapping is the preferred method for most NGS applications, particularly in clinical genetics (Flicek and Birney, 2009). However, de-novo assembly, on the other hand, is primarily limited to more focused tasks, such as correcting flaws in the reference genome and improving the detection of SV and other complicated rearrangements and newer findings (Ameur et al., 2018).

In the context of human clinical genetics, the third main phase of the NGS analysis pipeline addresses the essential issue of “making sense” or data interpretation, which requires finding the basic link between variant data and the observed phenotype in a patient. The tertiary analysis starts with variant annotation, which adds a fresh layer of data to predict the functional impact of all variants found during the variant calling procedure. Variant filtering, prioritisation, and data visualisation approaches are utilised after variant annotation. These procedures can be carried out utilising a number of software suites, which must be updated on a regular basis to reflect the most recent scientific findings, necessitating ongoing maintenance and development on the part of the developers. The generalised workflow of NGS is shown in Figure 2 .

Variant annotation is a crucial first step in the assessment of sequencing variants. As previously indicated (Scherer et al., 2007), variant calling generates a VCF file. Each line in such a file contains high-level information about a variant, such as genomic position, reference, and alternate bases, but no information biological implications. Variant annotation provides biological context for all discovered variants. Data annotation is performed automatically due to the large amount of NGS data. For variant annotation, several programmes are currently available, each of which uses distinct approaches and databases such as Sorting Intolerant from Tolerant (SIFT), (Ng and Henikoff, 2003), PolyPhen-2, (Adzhubei et al., 2010), Combined Annotation Dependent Depletion (CADD) (Kircher et al., 2014) and Condel (González-Pérez and López-Bigas, 2011), compute the impact scores for each variant based on various specifications, such as sequence homology, conservation of amino acid residues, evolutionary conservation, structure of protein, or statistical prediction based on known mutations, are integrated into such annotation tools. Furthermore, annotation can be used to search disease variant databases like ClinVar and HGMD for information on their clinical associations. Annotate Variation (ANNOVAR) (Yang and Wang, 2015) variant effect predictor (VEP) (McLaren et al., 2010), Single Nucleotide polymorphism effect (snpEff) (Cingolani et al., 2012), and SeattleSeq (Ng et al., 2009) are the most extensively used annotation tools among the many available. SNPs, INDELs, and Copy Number Variation (CNVs) can all be found using ANNOVAR, a command-line tool. It compares variants and explicates the functional consequence of variants on genes and other genomic components (Wang et al., 2010a). The overall number of variants obtained after analysis of a VCF file from WES may range between 30,000 and 50,000. Filtering algorithms are required to find the variant(s) responsible for a particular disorder. Some more examples in Table 1. As a result, it is strongly advised to eliminate false-positive calls and variant call errors when beginning the third level of NGS analysis, depending on quality parameters or prior knowledge of artefacts. The population frequency filter is one of the most widely used NGS filters. One of the filter based on allele frequency is minor allele frequency (MAF), which can sort variations into different categories: uncommon variants (MAF 0.5, usually picked for Mendelian illnesses), low frequency variants (minor allele frequency between 0.5 and 5%), and common variants (MAF >5%) (Consortium et al., 2010). It not only aids in better identifying disease alleles, but also in understanding population migrations, relationships, origins, admixtures, and population size changes, which may be useful in understanding various disease patterns (Stoneking and Krause, 2011). The most extensively utilised databases are the 1,000 genome project (Siva, 2008), Exome Aggregation Consortium (ExAC) (Lek et al., 2016), and the Genome Aggregation Database (gnomAD; http://gnomad.broadinstitute.org/). This filter, however, has limits and may result in incorrect exclusion.

Even though, functional annotation offers a significant information for filtering, the most critical question to answer, especially in the context of gene discovery, is whether a given variant or mutant gene the disease-causing gene? What is its frequency in different population sets studied globally? To solve this difficult issue, a new generation of tools is being created that, rather than just omitting information, rate variants and allow them to be prioritised. (MacArthur et al., 2012; Lelieveld et al., 2016; Harper, 2017). Various ways have been suggested e.g. PHIVE investigates the similarities between human illness phenotypes and those derived from animal model organism knockout experiments (Robinson et al., 2014). While other methods try to handle the problem in a novel way, by computing a lethal score (also known as burden score) for each gene using data from population variation databases (Eilbeck et al., 2017).

Phevor, which uses data from other relevant ontologies, such as gene ontology (GO), to advocate novel gene–disease connections, can also be employed for the identification of novel genes (Singleton et al., 2014). The fundamental purpose of these tools is to provide a small number of variants that can be validated using molecular techniques (Pereira et al., 2019a; Pereira et al., 2019b). VarSeq/VSClinical (Golden Helix), Ingenuity Variant Analysis (Qiagen), Alamut^® software (interactive biosoftware), and VarElect have all recently been developed commercial softwares for the elucidation and prioritisation of variants in a clinical context, to be used by clinicians, geneticists, and researchers (Stelzer et al., 2016). Apart from the tools that aid in variant analysis and elucidation, clinicians now have access to medical genetics firms like Invitae (https://www.invitae.com/en/) and CENTOGENE (https://www.centogene.com/) that provide a precise medical diagnosis.

5) Third generation sequencing technologies has the capability of sequencing single molecules with average read lengths of >10,000bp -100,000bp or even more. The advent of this technology has eliminated the requirement of amplification of DNA (PCR) and it provides real time results (Pereira et al., 2020). The third-generation sequencing services are provided by Pacific Biosciences (PacBio) that utilizes the single molecule real time (SMRT) platform and fluorescent nucleotide detection methodology. Oxford Nanopore Technologies (Minion) which utilizes the nanopore methodology where an ionic current passes through the flow cell and nucleotides bases are determined by the changes they produce in the current respectively when pass through the nanopores. (Xiao and Zhou, 2020).

The bioinformatic tools required to analyze the data obtained from the third-generation sequencing technologies needs to be more specific and error prone. Some tools are depicted in Table 2.

Limitations: Although Third generation sequencing technology is fast and provide real time result however still NGS are preferred as the error rate is less in NGS as compared to third generation sequencing which is ∼15%. Due to this high error rate, the technology can miss the detection of SNPs/point mutations and not best suited for mutational analyses. The methodology requires improvement. Moreover, there is need to develop more bioinformatic tools and algorithms for the downstream data analyses that is again a challenge for researchers for the time being (Ozsolak, 2012).

cDNA sequencing or RNA-seq when compared to other methods allows for more accurate mapping of reads and quantification at the transcript level. Differential expression analysis and identification of isoforms due to mRNA splicing, NGS of Small non-coding RNA as well as the discovery and characterisation of novel transcripts, are examples of high throughput applications (Marioni et al., 2008; Wang, 2009; Montgomery et al., 2010).

Small non-coding RNA NGS: A significant increase has been seen in the research community related to biomarkers which aids in the prediction, early detection and prevention of the disease. The biomarkers research helps the scientific and clinical community significantly in improving the clinical outcomes (Lopez et al., 2015). Non-coding RNAs (ncRNAs) have become the biomarker hotspot of the research interest in the field of disease identification and treatment. MicroRNAs (miRNAs) are the type of ncRNAs which are mostly explored for their potential biomarker role (Lopez et al., 2015). Till date ncRNA studies have been performed mainly by qRT-PCR, in situ hybridization, or microarray techniques. NGS has opened a new way to analyze/detect the RNA molecules present in the biological samples. NGS tenders several methodological advantages over other technologies like increased throughput, decreased RNA input, good consistency and quality of data, higher detection depth, analysis of all RNA populations, and discovery of novel molecules (Liu et al., 2021). A typical RNA-sequencing experiment consists of the following steps:

Thus all the above possibilities have allowed us to learn more about the genome’s organisation, the molecular constituents of cells and tissues, and the complexities of regulatory systems (Zhou et al., 2010; Sims et al., 2014). Many investigations, both fundamental and applied, have focused on mRNA splicing. Between the transcriptional and translational level, splicing occurs in every eukaryotic cell. Pre-mRNA transcripts may be variably spliced depending on location of tissue and/or stage of development, allowing multiple transcripts to be generated and hence distinct proteins to be made from the same gene (Burge et al., 1999; Nilsen, 2003). The divergence of splice site sequences from the prototypes has been linked to the generation of alternative transcripts. Furthermore, in most introns of higher eukaryotes, these extremely degraded motifs may be observed. Pseudo-exons are intronic sequences of standard exon size that outnumber real exons and are flanked by sequences that fit the exon’s 5′ and 3′ splicing signal requirements, but are never recognized as proper exons by the spliceosome. To distinguish true exons and splice sites from pseudo exons, splicing machinery must rely on auxiliary sequence features such as intronic and exonic cis-elements (Jacob and Gallinaro, 1989).

Exonic Splicing Enhancers (ESEs) are the most researched and well explored among them. They’re nucleotide sequences of short length that are primarily targeted by Serine/Argine-rich (SR) proteins, which then help to define exons (Blencowe, 2000). Exonic Splicing Silencers (ESSs), on the other hand, assist the spliceosome in neglecting pseudo exons and decoy splice sites. They serve as binding sites for exon exclusion-promoting proteins (mostly hnRNP proteins) (Zhu et al., 2001). Several bioinformatics approaches have been created and are now accessible to examine or predict splice signals (Zhang et al., 2005). One of the most essential bioinformatics tools is HSF (Human Splice Finder). For administration of data, designing of algorithm and online interface, HSF was built with the 4D package (4D S.A.). The HSF database was created with all human genes containing introns and exons. It was created using an Ensembl dataset that included about 22 000 genes and 46 000 transcripts from Homo sapiens. Because matrices and methods were specifically built for the human genome, the HSF database exclusively contains human genes (Flicek et al., 2008). HSF also has data taken from the Ensembl Variation Database (EVD), which can be used to investigate the impact of SNPs on splicing. A Perl script was written utilizing the Ensembl Perl API to allow HSF to access the EVD directly and get SNPs in human genes. Because matrices and methods were specifically built for the human genome, the HSF database exclusively contains human genes (Flicek et al., 2008).

On the other hand, Exonic splicing enhancers (ESEs) can be disrupted by nonsense, missense, and even translationally silent mutations, causing the splicing machinery to skip the mutant exon with significant consequences on gene structure. The frequency of mutations, whose major consequence is unusual splicing has been significantly underestimated because the effects of mutations are most often predicted purely based on information of genomic sequence (Cartegni et al., 2002). ESEs are found in both alternative and constitutive exons, where they serve as binding sites for Ser/Arg-rich proteins (SR proteins), a family of conserved splicing factors involved in a variety of splicing stages (Graveley, 2000). Through their RNA-binding domain, SR proteins promote exon definition by attracting spliceosomal components via protein–protein interactions facilitated by their RS domain and/or antagonizing the function of surrounding splicing silencers. Multiple categories of ESE consensus motifs have been described, and different SR proteins have varying substrate specificities (Graveley, 2000; Cartegni et al., 2002; Fairbrother et al., 2002). Using weight matrices for four different human SR proteins, ESE finder searches query sequences for potential ESEs. The matrices are based on frequency values produced from the alignment of winning sequences obtained through functional SELEX studies, corrected for the background nucleotide frequency of the initial SELEX library, which was created using chemical synthesis (Liu et al., 1998; Liu et al., 2000). The query sequences can be entered directly into the input box or submitted as a text file. Multiple sequences can be processed at the same time if they are preceded by a FASTA-format description line (starting with ‘>’). Despite the fact that ESEfinder is a tool for RNA analysis, it only accepts normal DNA nomenclature (A, C, G, and T, not U). Any character other than the letters A, C, G, and T, as well as spaces and paragraph breaks, will be ignored by the programme. Although both upper and lower case are acceptable, the output lines will be written in upper case. The user can choose from one to four matrices to be used at the same time. The result for each matrix is a series of 1 ntd incremented scores. Only the ‘hits’ or ‘high score motifs’ are displayed in the initial output window, Figure 3 which include the position of the first nucleotide, the motif match sequence, and the calculated score. When a score exceeds the threshold value set in the input page, it is deemed a high score.

By choosing the ‘custom’ button and entering the required value into the box, any score can be used as the cutoff threshold. As a result, ESEfinder may be used to identify potential ESEs, and the prime application is the accurate interpretation of the impact of disease-associated variants. It has been previously demonstrated that ESEs predicted by this matrix-based method cluster in places where natural enhancers have been empirically localized and are more common in exons than in introns (Cartegni et al., 2003).

CircRNAs: In contrast to messenger RNAs, circular RNAs (circRNAs) are physiologically active nucleic acid molecules that occur in closed loop RNA forms and do not have polyadenylated tails. CircRNAs are classified as non-coding RNA (ncRNA), yet some circRNAs have the ability to code for proteins. CircRNAs were originally discovered and identified in plant viroids in the 1970s, and then in the cytoplasm of eukaryotic cells in the 1980s. Due to the prevalence of linear RNAs, early development in this field was likely modest, and circRNAs were thought to be a consequence of RNA splicing. Recent advancements in next-generation sequencing and related bioinformatics technologies, on the other hand, have speed up research in humans, mice, nematodes, plants, and archaea have all been found to have these compounds (Chen et al., 2021). Various tools employed for the analysis of circRNAs are summarized in Table 3, below.

Understanding the molecular processes that mediate cellular physiology requires the identification, quantification, and characterization of a cell’s whole protein content (Schmidt et al., 2014; Jensen et al., 2006). A rapid advancement in proteomics has steered the researchers to organize the study of protein structure, function, relationships, and dynamics in space and time. The groundbreaking revelation that DNA contains all of the genetic instruction required to build an organism gave rise to molecular biology’s central dogma, which characterized a one-way flow of information from DNA to RNA to Proteins. This belief has been debunked by recent discoveries. Epigenetic markings, alternative splicing, non-coding RNAs (including microRNAs), protein–protein interaction (PPI) networks, and post-translational modifications (PTMs) are only a few examples of how genotype and phenotype are not solely determined by information on the genome (Nagaraj et al., 2011; Beck et al., 2011; Baker, 2012). Proteomics is the global study of proteins, which are the key functional entities in the cell. This analysis is arguably the most important level of information required to understand how cells work. When compared to data collection at the genomic and transcriptomic levels, the proteomic data acquisition has proven difficult. Global protein analysis is a difficult analytical task, in part because amino acids, the building blocks of proteins, have such a wide range of physicochemical properties. Furthermore, in comparison to the genome, the proteome is enriched by alternative splicing and a wide range of protein modifications and degradation, and the complexity is heightened by the interconnectivity of proteins into complexes and signaling networks that are highly divergent in time and space Figure 4 (Cox and Mann, 2011). A decade ago, sequencing and identifying a single protein was a big problem; however, today’s high-throughput technology allows for the identification and quantification of essentially all expressed proteins in a single experiment. Similarly, 10 years ago, MS-based phosphoproteomics could only identify a few hundred phosphosites, whereas currently more than 30,000 phosphosites can be quantitatively monitored. This current method is referred to as “next-generation proteomics” to reflect its ability to characterize practically the whole proteome as a result of advancement in technology. Proteomics technologies, particularly MS-based Protein identification has advanced tremendously in recent years as a result of cumulative technological breakthroughs in instrumentation, sample preparation and computational analysis (Ficarro et al., 2002; Lemeer and Heck, 2009; Lundby et al., 2012).

Proteomics using mass spectrometry (MS) generates a large quantity of information about the expression, post-translational modifications (PTMs), and interactions among thousands of proteins. The obtained data must be supplied to the scientific community in a format that is both suitable and curated, as well as retrievable and interpretable. Proteomics data will be made freely available to the public, ensuring that quality standards are maintained in the area. The long-term storage of unprocessed raw data is a first level of distribution for proteomics data. Understanding the proteome’s complex and dynamic interactions necessitates the creation of physical interaction charts.

Proteins frequently interact with one another in stable or transient multi-protein complexes of varying composition, with the human interactome containing an estimated 130,000 binary interactions, the majority of which have yet to be mapped. Proteins can also interact with other molecules like RNA, DNA and metabolites. These complexes play crucial roles in regulatory processes, signalling cascades, cellular functions, and their inability to interact can result in their function being lost (Altelaar et al., 2012; Ma and Johnson, 2012). Tranche is one of the few public repositories that can manage this type of data at the moment, and it is based on an encrypted peer-to-peer system that stores data in numerous servers across the world. Raw data, on the other hand, is in a closed format, which makes it difficult to share. As a result, attempts are being undertaken to standardise formats that preserve all necessary information (Smith et al., 2011). The European Bioinformatics Institute’s PRIDE database exhibits this determination, as it enables the for the storage of both conventional MS data formats (XML) and associated peptide and protein identifications. Furthermore, including additional data (such as species, fragmentation procedures, and proteases) allows for a global meta-analysis of proteomic data sets (Perez-Riverol et al., 2019).

Moreover, Protein sequence alignment compares two or more than two sequences and aids in the identification of homologous regions, visualizing the relationship among sequences with respect to evolution and structure. It plays a crucial role in bioinformatics and helps in the query and construction of databases, prediction of protein’s primary, secondary and tertiary structure and biological function and many more. Many platforms are developed to analyse the sequence alignment. Some of them are PROSITE, Pfam, BLAST, FASTA, Clustal omega, T-Coffee, MUSCA, ALIGN, DIALIGN, ProbCons, HMMER3 phmmer and many more (Pruess and Apweiler, 2003; Sievers et al., 2011; Singh et al., 2016a).

Protein structure prediction can be done using the ProtParam tool from ExPasy (Expert Protein analysis system) (Gasteiger et al., 2005). It helps in the primary structure prediction of protein and aids in the computation of physicochemical properties of a given protein. The parameters that can be computed include molecular weight, amino acid and atomic composition, isoelectric point, estimated half-life, grand average of hydropathicity (GRAVY) and more. To predict the secondary structure, many tools have been developed till now including Chow-Fasman algorithim—a statistical approach which is based on calculation of statistical propensities of each residuum to form an α-helix or β-strand, GOR, Jpred, etc. Similarly, for tertiary protein structure prediction, PHYRE2 (Protein Homology/analogY Recognition Engine) (Kelley et al., 2015) and I-TASSER are available (Yang et al., 2015).

Apart from above mentioned software suits, there are other tools which are helpful in addressing protein analysis. Some of them are mentioned in Table 4.

To study the post-translational modifications, tools like GlycoMod (Cooper et al., 2001), NetPhos (Trost and Kusalik, 2011), NetPicoRNA (Smits et al., 2013), FindMod (Gasteiger et al., 2003), ScanProsite (De Castro et al., 2006) and others are available online. For protein interaction analyses STRING can be used (Szklarczyk et al., 2021). To visualize the 3-D structure of proteins, tools like Pymol and Jmol can be used. Pymol is also used to visualize the protein-ligand docking, binding site prediction, protein interactions and others (DeLano, 2002; Herráez, 2006).

The identification of protein biomarkers with prognostic or diagnostic significance is one of the most difficult applications of proteomics right now Figure 5 .

As previously mentioned, recent technical advancements have resulted in the development of comprehensive pipelines that incorporate the discovery and validation phases, allowing plasma biomarkers to be identified for many diseases (Addona et al., 2011; Whiteaker et al., 2011). Despite the introduction of some successful biomarkers for clinical application, many (if not most) claimed biomarkers have weak reliability or lack rigorous confirmation, leading to scepticism among clinicians. The lack of proper controls in the discovery phase, the use of appropriate statistical tools for biomarker definition, and the need for independent validation steps in large patient cohorts to certify the legitimacy of the biomarker unambiguously are the primary flaws in many biomarker studies; such flaws lead to claimed biomarkers that are rarely directly related to disease biology (Poste, 2011).

The introduction of massively parallel sequencing has changed genetics and genomics research forever because of its widespread adoption and numerous applications, massively parallel sequencing is projected to play a vital role in the medical industry in the next years. It is worth noting that NGS as a research tool faces major challenges in terms of manufacturing, data management and downstream analysis.

➢ Thus, in the past decade, rapid advancements in high-throughput intervention, backed by lower costs, have opened up new pathways for interrogating a biological system at several regulatory levels, while also providing us with an unprecedented picture. Integrating more genomic/proteome/transcriptome/metabolome/epigenome data with relevant information obtained at other levels, such as genomes, transcriptomes, epigenomics and metabolomics is still a difficulty.