Exploring KEGG Pathway Database: A Bioinformatics Guide to Metabolic Pathway Analysis

Bioinformatics is an interdisciplinary field combining computer science, biology, information technology, and mathematics to analyze biological data. It encompasses physical map construction, gene discovery, molecular function analysis, and 3D structure interpretation. Biological databases are organized collections of data enabling researchers to locate, add, and modify information. Data exists in two forms: sequences (one-dimensional, including DNA, RNA, and protein sequences) and structures (three-dimensional spatial arrangements). The Protein Data Bank (PDB), established in 1956, serves as the first protein structure database and now contains over 83,000 structures. Major public sequence databases include GenBank, EMBL, and DDBJ, which store raw nucleic acid sequence data. Secondary databases like Swiss-Prot provide high-level annotation with minimal redundancy.

Biological databases are structured electronic repositories that organize, store, and retrieve biological data such as nucleotide sequences, protein sequences, macromolecule structures, gene expression profiles, and literature, characterized by two essential features: non-redundancy (minimizing duplicate entries) and data sharing protocols (allowing scientists to examine, verify, and update information), with common types including sequence databases (for genes and genomes), structural databases (for protein topology), literature databases (like PubMed Central), gene expression databases (such as GEO and SAGEmap), and chemical databases (like PubChem).

A database in bioinformatics is a library-like structure designed to store and organize biological data. After DNA or RNA sequencing, the resulting sequence data is uploaded to specialized databases. These databases can be categorized into different types: some contain only nucleotide sequences, others contain only protein sequences, and some may include both. The purpose of these databases is to allow researchers to compare new sequences against existing ones to identify similarities and potential functions.

Bioinformatics databases: PDB (RCSB, PDBE, PDBj) stores 3D structures with four-letter codes; CATH classifies structures into classes (alpha, beta, alpha+beta, few) with architecture, topology, and homologous superfamily; Gene Ontology (GO) provides controlled vocabulary with three domains (biological process, molecular function, cellular location) and hierarchical IDs; KEGG organizes biological pathways. Database IDs link across resources. Serious bioinformatics requires Unix command-line tools and programming skills beyond web interfaces.

A biological database is an organized collection of related biological data (nucleotide sequences, protein structures, etc.) that can be easily stored, accessed, and managed; the ten main types include bibliographic databases (PubMed for biomedical literature), sequence databases (GenBank for nucleotide sequences), structure databases (PDB for 3D protein/nucleic acid structures), metabolic databases (KEGG for biological pathways), model organism databases (EcoCyc for E. coli), enzyme databases (BRENDA for enzyme information), disease databases (OMIM for human genetic disorders), chemical databases (PubChem for small molecules), gene expression databases (GEO for microarray data), and taxonomic databases (Catalogue of Life for species classification).

Enzymes are classified into six main classes based on reaction type: (1) Oxidoreductases - catalyze oxidation-reduction reactions, (2) Transferases - transfer functional groups between molecules, (3) Hydrolases - catalyze hydrolysis reactions (bond breaking with water), (4) Lyases - catalyze bond breaking without hydrolysis, often forming double bonds, (5) Isomerases - catalyze isomerization reactions, (6) Ligases - catalyze bond formation between molecules, often using ATP. Each enzyme has a unique four-digit EC number indicating its classification.

The Enzyme Commission Number (EC number) is a four-digit number system used to classify enzymes. The first digit represents the class of enzyme (1 for Oxidoreductases, 2 for Transferases, 3 for Hydrolases, 4 for Lyases, 5 for Isomerases, 6 for Ligases, and 7 for Translocases). The second digit represents the sub-class, the third digit represents the sub-sub-class, and the fourth digit represents the individual enzyme. For example, SGOT (Serum Glutamic-Oxaloacetic Transaminase) has EC number 2.6.1.1, indicating it belongs to the Transferase class (2), sub-class 6 (transfer of nitrogenous compounds), sub-sub-class 1 (transfer of amino groups), and is the first enzyme in this category.

The International Union of Biochemistry and Molecular Biologists (IUBMB) classifies enzymes into six main classes (now seven as of 2018), each assigned a four-digit enzyme commission number. Class 1 (Oxidoreductases) catalyze oxidation-reduction reactions (e.g., alcohol dehydrogenase). Class 2 (Transferases) transfer functional groups (e.g., transaminases). Class 3 (Hydrolases) catalyze hydrolysis reactions using water (e.g., trypsin). Class 4 (Lyases) remove groups to form double bonds (e.g., aldolase). Class 5 (Isomerases) catalyze isomerization reactions (e.g., retinal isomerase). Class 6 (Ligases) join compounds together (e.g., DNA ligase). Class 7 (Translocases, added 2018) transfer ions across membranes (e.g., ATPase).

The International Union of Biochemistry and Molecular Biology (IUBMB) classifies enzymes into six major classes based on the type of reaction they catalyze, each assigned a unique EC (Enzyme Commission) number: (1) Oxidoreductases (EC1) catalyze oxidation-reduction reactions involving electron transfer, such as alcohol dehydrogenase and lactate dehydrogenase; (2) Transferases (EC2) transfer functional groups between molecules, like hexokinase transferring phosphate groups; (3) Hydrolases (EC3) catalyze hydrolysis reactions breaking bonds using water, exemplified by amylases and lipases; (4) Lyases (EC4) catalyze addition or removal of groups to form double bonds, including aldolase and carbonic anhydrase; (5) Isomerases (EC5) catalyze isomerization changing molecular arrangements within a molecule, such as phosphoglucose isomerase; (6) Ligases (EC6) join molecules together using ATP hydrolysis, like DNA ligase and glutamine synthetase.

The IUBMB established a standardized classification system for enzymes using EC numbers (Enzyme Commission) following the format EC X.X.X.X, where each digit represents hierarchical classification levels. Enzymes are classified into six main classes based on reaction type: (1) Oxidoreductases (EC 1) catalyze redox reactions; (2) Transferases (EC 2) transfer functional groups; (3) Hydrolases (EC 3) break bonds using water; (4) Lyases (EC 4) form/break bonds without ATP/water; (5) Isomerases (EC 5) convert molecules to isomers; (6) Ligases (EC 6) join molecules using ATP energy. This systematic approach enables precise identification and comparison of enzymes across scientific literature.

Systems biology emerged in the mid-1990s as a genome-enabled science following genome sequencing. The four-step paradigm involves: (1) understanding cellular components, (2) reconstructing interactions into networks, (3) converting to computational models, and (4) computing phenotypic functions. Biological networks are based on chemical transformations involving covalent bond changes or molecular associations, consisting of nodes (chemical compounds) and links (transformations). Three main network types are metabolism (small molecule transformations), transcriptional regulation (DNA-binding proteins controlling gene expression), and signaling networks (cell-to-cell communication). Mathematical models serve five purposes: organizing diverse information, quantitatively relating variables, discovering new strategies, understanding qualitative features of complex processes, and the principle that 'what I cannot create, I do not understand.' Biology operates across multiple scales: size scales from nanometers (molecules) to microns (cells), time scales from milliseconds (diffusion) to hours (bacterial doubling), and abundance scales from single molecules per cell to millimolar concentrations. Networks have steady states (unchanging) and dynamic states (changing). Multiscale analysis shows that as different events relax on different time scales, dynamic dimensionality shrinks, reducing independent variables and enabling variable aggregation.

System biology requires thinking in networks rather than isolated components. Living organisms exist at multiple interconnected levels: genes influence each other, proteins form networks, 37 trillion cells interact, and humans exist within social and environmental networks. This complexity makes traditional reductionist approaches insufficient. System biology uses mathematics to model these networks, enabling predictions about biological processes. This approach has already successfully explained disease spread patterns and is now being applied to understand individual health at unprecedented levels of detail.

Like a car's components, biological parts (genes, DNA, proteins) interact in networks to create properties more complex than individual parts. A simple yeast cell diagram shows genes represented as balls connected by lines representing interactions. Some genes are highly connected while others are loosely connected, forming subsets that carry out specific cellular functions. Because components are organized into networks within single cells, alteration of one component through mutation can affect many other parts of the network, representing a different way of understanding disease.

Systems biology is an interdisciplinary approach that studies how biological components interact within complex networks. Rather than studying individual parts in isolation, systems biology examines how components like genes, proteins, and cells work together to produce emergent behaviors. This approach helps explain phenomena like circadian rhythms and ecosystem dynamics by analyzing the interactions between multiple components.

Systems biology is an approach that studies collections of structures and how they interact with one another to make the whole unit work. Instead of studying a single chloroplast, systems biology might study an entire cell to see how all parts work together, then how cells form a leaf, how organs form a tree, and how trees form an ecosystem. This approach works in the opposite direction of reductionism, moving from small individual parts to big systems.