**Abstract**

Metabolomics utilizes analytical profiling technique for measuring and comparing large numbers of metabolites produced in the biological fluids. Traditional process of drug development is not sufficient enough to understand the proper biochemical processes within the targets which may finally lead to the failure. Metabolomics can be very useful to overcome such failure as it involves in the detailed profiling and understanding of the biochemical processes which helps in identification of target engagement (TE) markers as well as predicting mode of action (MOA). Currently pharmaceutical companies are utilizing this approach in the early drug development stage to combat failure. This chapter will mainly highlight the advantages of this concept over traditional concept of drug development along with recent developments of it.

**Keywords:** metabolomics, drug, disease, drug discovery and development, analytical approaches

## **1. Introduction**

Pharmaceutical company that wants to thrive in this highly competitive field of introducing novel therapeutic agents for a variety of ailments at an extremely high pace, must devote a significant number of resources to the process of drug discovery. To create newer, more potent, and safer medications, they must make strides and look at different possibilities. The drug development process scenario has seen significant change since the turn of the twenty-first century [1].

From the initial step of target validation through clinical trials to clinical practice, the drug development process is a drawn-out, painful, and incredibly expensive one. The net output of such a method has also been subpar and frequently results from small improvements in currently available therapeutic agents. The current paradigm for drug development asks for the discovery of particular molecular targets that can be used to create highly powerful and selective inhibitors with little off-target activity. These substances may be synthetic small compounds that need medicinal chemistry optimization, or they may be natural products and their synthetic derivatives.

A systems biochemical understanding of the disease, the therapeutic agents' pharmacological properties (i.e., absorption, distribution, metabolism, excretion, and toxicity, or ADMET), and their functional impact on the human body both ontarget and off-target would logically be necessary for the entire process, from target discovery to target validation to clinical testing and ultimately clinical adoption.

System biochemistry refers to the "global biochemical networks and molecular regulations". This is a challenging task for drug discovery, development, and its application using conventional methods, as is the case with all systems techniques. The effective and prosperous commercialization of promising therapeutic medicines is fundamentally hampered by the absence of systems biochemical methods and, consequently, functional understanding in the past. A systems biochemical understanding of the human body can now be envisioned for the first time due to the advancements in genomes, functional genomics, proteomics, and now metabolomics. Once created, this will hasten the knowledge of disease mechanisms and the creation of therapeutics at a rate never before seen. Comprehension of the biology of disease requires a thorough understanding of metabolism in human disorders. The metabolome is an essential component of molecular homeostasis regulation. Metabolomics, or the investigation of the metabolome, is currently being utilized for treating a wide range of disorders, mostly through metabolite profiling for biomarker discovery [2].

In recent years, the rapidly developing discipline of metabolomics has taken on a significant role due to its numerous applications in the area of drug discovery and development. Metabolomics has recently made significant strides, and it may now be used as a key tool in the process of finding and developing new drugs.

### **2. Metabolomics and its evolution**

Dr. David Wishart's Human Metabolome Project has completed a year of study on the human metabolome, which contains 2500 metabolites, 1200 medicines, and 3500 dietary components. The \$7.5 million research recruited 53 scientists and archived findings on the Human Metabolome Database. The study employs cutting-edge techniques such as NMR spectroscopy, mass spectrometry, multidimensional chromatography, and machine learning to profile metabolites without bias and characterize metabolite interactions using multivariate methodologies [3]. Professor Jeremy Nicholson first put up the idea of metabolomics in 1999. Using a similar principle to genomics and proteomics, metabolomics is a method to quantitatively analyze all metabolites produced by organisms to determine their link to pathological and physiological changes [4, 5].

The substrates and end products of metabolism, known as metabolites, fuel vital cellular processes such as energy production and storage, signal transduction, and apoptosis. Metabolites can come from bacteria, xenobiotics, food, and other foreign sources as well as to being produced naturally by the host organism [6]. Metabolomics is the study of the organism's internal store of non-proteinaceous small molecules. A thorough examination of the entire metabolome (the sum of all the low molecular weight compounds that are present in cells during a specific physiological condition. It alludes to the list of molecules found in a particular organism) under a specific set of circumstances is called metabolomics. Metabolomics is the only technique that can quantify interactions between the genome, proteome, and the biological "wild card" known as the outside environment. The emerging field of genomic science is metabolomics. The metabolomic society correlates with other post-genomic sciences; ideally, metabolomic data sets will be merged with its other omic sciences to provide comprehensive views

*Metabolomics: Special Emphasis on Basic Drug Discovery and Development DOI: http://dx.doi.org/10.5772/intechopen.112969*

into the molecular processes of system biology. However, unlike genomics or proteomics, which concentrate on characterizing huge macromolecules (DNA, RNA, and proteins), metabolomics concentrates on characterizing the small molecule, catabolic, and metabolic products resulting from the interactions of these large molecule [ 7 ].

### **3. Metabolomic study design**

 Metabolomics experiment involves Experimental design, sample collection and preparation, sample analysis, data processing, and interpretation ( **Figure 1** ) .

#### **3.1 Experimental design**

 A proper experimental design is crucial for accurate interpretation of data, including sufficient subjects, matching covariates, proper sample collection, and appropriate data analysis techniques [ 9 ].

#### *3.1.1 Sample collection and preparation*

 Metabolic profiling analyses metabolites in both vivo and vitro samples. Metabolomics can analyze various biological materials, including biofluids, cells, tissue, and feces [ 10 ]. Standardized procedures for sample collection and storage improve

#### **Figure 1.**

 *Metabolomic workflow [ 8 ].* **Note:** *NMR (nuclear magnetic resonance), MS (mass spectroscopy), FTIR (Fourier transform infrared spectroscopy).* 

quality and reproducibility. Factors like fed vs. fasting state, medications, blood collection time and processing time, addition of additives and potential sample hemolysis considerations can affect metabolites, leading to false-positive or false-negative results. Sample preparation depends on the analytical approaches to be utilized for analysis, such as NMR approaches requiring less preparation as a result it does not affect the sample much whereas mass spectrometry-based approaches requiring sample extraction by using different solvents. Metabolomics studies examine metabolites at specific time points, enabling more dynamic assessments of specific metabolic pathways. Recent development by introducing metabolite with isotope (13C) allows dynamic assessment of metabolic pathways, determining intracellular molecule sources, metabolite fate, flux, and cellular redox balance, providing details information which are not available in case of steady-state metabolomics experiments [10–14].

Sample pretreatment plays an important role. Pretreatment separation modalities include gas chromatography for gas phase separation of molecules useful in the analysis of traces of volatile compounds in samples, high performance liquid chromatography allows high pressure elution that leads to increase in the chromatographic separation of the samples makes it more versatile separation technique, capillary electrophoresis works on the principle of electrokinetic separation useful in the separation of small inorganic ions to larger proteins. These techniques offer advantages in characterizing specific aspects of a metabolome [15].

#### **3.2 Detection methods**

Targeted and untargeted metabolomic analyses are the two main categories that can be used, in theory. Targeted analysis would concentrate on a certain number of identified compounds. Untargeted metabolomics, also known as discovery metabolomics, tries to gather all the metabolomic data in a sample, while In the latter, features of relevance are identified after being filtered using various uni and multi-variate statistical techniques following data capture [16].

For the isolation and quantification of metabolome components, a wide range of targeted and untargeted approaches have already been documented in the literature. It was discovered, however, that no one analytical platform is able to collect complete metabolomics data in a single run (**Figure 2**) [17].

Metabolic profiling manly based upon the two specialized analytical techniques viz. NMR spectroscopy and MS. These techniques are efficient enough to identify and quantify wide range of molecules requiring small amount of sample. NMR spectroscopy based on the frequency pattern resulting from the interaction of nuclei of the molecule with the electromagnetic field. This pattern can give the information such as structure of the molecule, its motion, and chemical environment [18]. The identification and measurement of metabolites utilizing NMR techniques, such as proton NMR, 13C NMR, 19F NMR, and 31P NMR spectroscopy, have improved recently. The development of cryoprobes and microprobes, which have decreased the detection limit by a factor of around 3 to 5, is noteworthy. This approach is further enhanced by using two-dimensional total correlation spectroscopy (2D TOCSY) for the confirmation of assigned peaks. Other examples of two-dimensional NMR that have been used to enhance NMR-based data acquisition and metabolite structure analysis include Nuclear Overhauser effect spectroscopy (NOESY), heteronuclear single quantum coherence (HSQC), exchange spectroscopy (ES), and J spectroscopy (JS). These techniques provide better information than one-dimensional NMR, particularly for small-molecule metabolites. Furthermore, better outcome in the metabolome

*Metabolomics: Special Emphasis on Basic Drug Discovery and Development DOI: http://dx.doi.org/10.5772/intechopen.112969*

#### **Figure 2.**

*Major analytical platforms for metabolomic studies in human and animal samples.* **Note:** *HPLC (high performance liquid chromatography), GC (gas chromatography), CE (capillary electrophoresis).*

analysis can be achieved by the combination of 2D NMRs such as NOESY and HSQC and TOQCY and HSQC, by combining with MS with NMR or by 3DNMR [8]. NMR spectroscopy is rapid, non-destructive and gives reproducible results which makes its most reliable. Furthermore, samples that have undergone NMR analysis may undergo subsequent MS analysis. This technology's shortcomings are its lack of sensitivity and high user obtaining the necessary tools comes at a significant initial start-up cost and requires special training [19]. In contrast, the destructive analytical method known as mass spectrometry relies on the production of gas phase ions that are then distinguished by their charge/mass ratio. The number of ions for each mass/charge ratio is then calculated once the ions have reached the detector. In order to determine the molecular identities of the constituents, this is processed and compared against accessible mass spectral databases. MS is a very sensitive sample analysis technique that may be applied to both targeted and non-targeted analyses. However, the experimental setup and the instrument parameters have a significant impact on the detection's sensitivity and accuracy [18, 19].

The MS methodology is typically combined with chromatographic methods that have variable degrees of sensitivity, such as liquid chromatography (LC-MS), particularly high-performance liquid chromatography (HPLC- MS), and gas chromatography (GC-MS). Capillary electrophoresis and the MS method can also be combined for better outcome (CE-MS) [20].

A part from NMR and MS, FTIR is another tool which can be utilized successfully for metabolomic study (**Table 1**) [21].

#### **3.3 Data analysis**

This phase entails identifying metabolites and figuring out their relative abundance. The platform and method (targeted vs. untargeted) affects how metabolite identification occurs [23].

Standards are typically run in targeted studies and metabolite identification is less uncertain [24]. Whereas metabolite identification in the untargeted technique


#### **Table 1.**

*Comparison of commonly employed analytical techniques.*

is uncertain comparatively, the spectra's are analyzed using either proprietary or publicly accessible software [25].

Additionally, the data should be checked for anomalies and samples or metabolites with a large number of missing values. Samples or metabolites that do not adhere to quality control standards should be eliminated at this stage. These actions result in the acquisition of a set of robustly measured metabolite [26, 27]. Rapid and precise statistical tools are required to handle the complexity and volume of the enormous amount of created data. For data analysis, several metabolomic features may be employed as the input. Spectral bin areas, metabolite concentrations, and spectral peak areas are some of these [28].

Numerous univariate and multivariate statistical methods can be used, focusing on data pre- and post-processing tasks such peak fitting, noise reduction, run order drift correction, and signal extraction/peak recognition. They are collectively referred to as chemometric approaches. The metabolomic characteristics are independently analyzed using univariate techniques [29]. Due to the fact that they use more widely accepted and understood statistical techniques, they are frequently simpler to interpret. However, the presence of interactions between various metabolic features is not taken into account in this approach. Confounding factors like gender, nutrition, or BMI are not taken into account. This raises the likelihood of receiving inaccurate results. Unlike univariate analysis, multivariate analysis takes into account all imputed metabolomic variables and attempts to uncover connections between them. These methods are divided into two categories: supervised methods and unsupervised methods. The most prevalent unsupervised method is principal component analysis, which is capable of detecting data patterns with biological variables. Supervised approaches find patterns within variables of interest while ignoring other sources of variation. Partial least squares regression analysis is the most commonly used supervised statistical procedure [30].

There are a number of software tools available for doing metabolite set enrichment analysis and visualizing the results. Metaboanalyst (www.metaboanalyst.ca)

#### *Metabolomics: Special Emphasis on Basic Drug Discovery and Development DOI: http://dx.doi.org/10.5772/intechopen.112969*

is a website for metabolomics data analysis, which includes many tools for pathway enrichment analysis [31]. Metlin (https://metlin.scripps.edu), a massive database of metabolites with their MS-derived ions that serve for creative pathway analysis, is an online platform that allows extensive analyses and interpretation of omics data [9]. Metabox (free at http://kwanjeeraw.github.io/metabox/) (retrieved on January 10, 2020) under the GPL-3 license) [32] and MetaboAnalyst (http://www.metaboanalyst. ca/MetaboAnalyst [accessed: January 10, 2020]) [22]. Others such as SECIMTools, Meta XCMS, XCMS, XCMS2, MetAlign, MZmine for MS data processing, and MetDAT for statistical analysis and pathway visualization are among the tools available [8].
