<h5 id="91-the-utility-of-basic-mathematical-and-statistical-concepts-to">9.1 The Utility of Basic Mathematical and Statistical Concepts to</h5> <ul> <li>Biology requires a fundamental understanding of probability.</li> <li>Statistical knowledge is crucial in designing experiments, choosing sample sizes, and repeating tests for confidence in results.</li> <li>Familiarity with statistical tests and measures of significance is necessary to determine the validity of results.</li> <li>Advanced computing, analysis, and visualization can be achieved through frameworks like MATLAB and R.</li> <li>The choice of statistical analysis is critical in obtaining accurate results; false assumptions and unbalanced models can lead to false positives.</li> <li>Caution must be exercised in applying mathematical models and statistical tests to ensure validity and interdisciplinary advancements in biology.</li> </ul> <p>======</p> <h5 id="92-introduction">9.2 Introduction</h5> <ul> <li>Bioinformatics is an interdisciplinary field that combines biology with computational, mathematical, statistical, and engineering approaches.</li> <li>It involves the storage, retrieval, analysis, and interpretation of biological data using computer-based software and tools.</li> <li>Often used interchangeably with terms such as ‘computational biology,’ ‘mathematical biology,’ ‘quantitative biology,’ and ‘bio-statistics.’</li> <li>The definition of bioinformatics can vary among experts and has evolved over time.</li> <li>Fig. 9.2 illustrates the interdisciplinary nature of bioinformatics, showing its intersection with computer science, mathematics, engineering, and statistics.</li> </ul> <p>======</p> <h5 id="921-historical-perspective">9.2.1. Historical perspective</h5> <ul> <li>Bioinformatics is a field that aids in discovery through data mining and generation of new hypotheses via modeling or analysis of molecular data.</li> <li>It relies on preexisting nucleotide and protein data from databases or newly created data from high-throughput instruments.</li> <li>The National Center for Biotechnology Information (NCBI) in the USA is a resource for bioinformatics tools and services, housing nucleotide and bibliographic databases like GenBank.</li> <li>The term bioinformatics started appearing in the literature in 1991, gaining wide acceptance after the launch of the human genome project.</li> <li>The primary concern in bioinformatics is managing sequence and structural data in databases and mining data to get biological meanings.</li> <li>Bioinformatics uses databases based on secondary and tertiary levels of information and continually updates to address data discrepancies and errors.</li> </ul> <p>======</p> <h5 id="922-types-of-experimental-technologies-for-analysing-biomolecules">9.2.2. Types of experimental technologies for analysing biomolecules</h5> <ul> <li>PCR (Polymerase Chain Reaction) is a technique used to amplify a region of interest in DNA.</li> <li>Real-Time-PCR/qPCR (quantitative PCR) is used to detect RNA expression by amplifying and quantifying cDNA synthesized from RNA.</li> <li>Next-generation sequencing is used to sequence genes/genomes and RNA by determining the order of nucleotides in DNA or RNA.</li> <li>Gel electrophoresis is used to separate fragments of DNA, RNA, and proteins based on their size and charge.</li> <li>HPLC (High-performance Liquid Chromatography) is used to separate, identify and quantify metabolites.</li> <li>MS (Mass spectrometry) is used to fragment, measure isotopic composition and determine mass of DNA, proteins, metabolites.</li> <li>EM (Electron microscope) is used to determine the structure and sequence of DNA, RNA or protein.</li> </ul> <p>======</p> <h5 id="923-types-of-molecular-data">9.2.3 Types of molecular data</h5> <ul> <li>There are various types of molecular data produced by different technologies.</li> <li>Two commonly used DNA data formats are FASTQ and FASTA.</li> <li>FASTQ format represents sequence data along with quality scores.</li> <li>FASTA format represents sequence data without quality scores.</li> <li>Details about these formats are provided later in the chapter (Genome Informatics).</li> <li>No equations or formulae are provided in the text.</li> </ul> <p>======</p> <h5 id="924-commonly-used-analytical-and-statistical-workflows">9.2.4 Commonly used analytical and statistical workflows</h5> <ul> <li>Biological knowledge is essential for the implementation and interpretation of results from various biological analysis tools.</li> <li>Some commonly used bioinformatics tools and their functions are: <ul> <li>Homology search: BLAST - blastn, blastp</li> <li>Sequence alignment: CLUSTAL, MAFFT, MUSCLE</li> <li>Phylogenetics: PHYLIP, PAUP</li> <li>Gene prediction: GlimmerHMM, GenScan</li> <li>Functional homology search: HMMER</li> <li>RNA structure: mfold, sFold, uniFold</li> <li>Regulatory region analysis: MatInspector, BEARR, RSAT</li> <li>Protein structure: Phyre2, Jpred</li> </ul> </li> <li>Statistical packages commonly used in bioinformatics are: <ul> <li>Statistical Package for the Social Sciences (SPSS)</li> <li>Statistical Analysis System (SAS)</li> <li>R</li> <li>Microsoft Excel</li> </ul> </li> <li>Functional interpretations of gene and protein level findings can be made using biological tests like loss- or gain of function assays, gene knockouts, and gene editing. Computational tools can be used to infer affected networks and pathways, assigning functional significance to a gene and its protein product.</li> </ul> <p>======</p> <h5 id="93-biological-databases">9.3 Biological Databases</h5> <ul> <li>A biological database is a collection of organized, structured, and searchable biological data.</li> <li>It can contain various types of information such as DNA sequence, mutation in a gene, protein sequence, 3D structures of proteins, and functional interactions.</li> <li>Biological databases can be accessed easily, have user-friendly interfaces, provide support staff, lack errors, are cross-referenced, and are continually updated.</li> <li>The two main types of databases are relational and non-relational, managed by database management systems (DBMS) and the structured query language (SQL).</li> <li>Non-relational or NoSQL databases are used for large sets of distributed and unstructured data.</li> </ul> <p>======</p> <h5 id="931-what-is-the-need-for-a-biological-database">9.3.1 What is the need for a biological database?</h5> <ul> <li>The need for biological databases arises from the complexity and vastness of biological data, which is challenging to search physically.</li> <li>Databases store biological information in a machine-readable format, allowing easy and accurate searches using user interfaces.</li> <li>Examples of commonly used biological databases include GenBank, PDB, UniProt, PubMed, KEGG, and OMIM, each storing specific types of biological information.</li> <li>The stored data can include DNA sequences, 3D protein structures, protein sequences, biomedical literature, biological pathways, diseases, drugs, and genetic disorders.</li> <li>Biological databases are essential for managing the exponential growth of genome data and facilitating research in various biological fields.</li> </ul> <p>======</p> <h5 id="932-data-visualisation">9.3.2 Data visualisation</h5> <ul> <li>The UCSC Genome Browser is an online interactive site that provides visualization of macro- and micro-level genome information for various vertebrate and invertebrate species.</li> <li>KEGG, Biocarta, and Reactome are resources that offer pathway visualization.</li> <li>CIRCOS is a tool for visualizing data in a circular layout, which can be useful for genomic data.</li> <li>Excel can generate histograms, scatter plots, bubble charts, and heat maps, while R is a software environment for statistical computing and generating graphics.</li> <li>D3.js is a JavaScript library for producing dynamic, interactive data visualizations in web browsers, and Phinch is an interactive, exploratory framework for visualizing biological data.</li> <li>The Integrative Genomics Viewer (IGV) is a tool for interactive exploration of large, integrated genomic datasets.</li> </ul> <p>======</p> <h5 id="94-genome-informatics">9.4 Genome Informatics</h5> <ul> <li>Genome is the complete set of DNA of an organism, including genes and intergenic regions. Genomics studies its structure, function, evolution, mapping, and modification.</li> <li>Genome informatics applies bioinformatics tools to process genome-wide assays data, linking it to function.</li> <li>High-throughput methods generate large amounts of genome data using various instruments, providing information on DNA/RNA nucleotide sequence, variations, expression, protein binding, and methylation.</li> <li>The Broad Institute of MIT and Harvard produced 24 TB of genome data per day in early 2018, equivalent to nearly 5000 personal computers worth of data a year.</li> <li>Understanding genome data’s complexity, patterns, and meaning requires computer science, information technology, quantitative methods, analytics, and statistics.</li> </ul> <p>======</p> <h5 id="941-human-genome-project">9.4.1 Human Genome Project</h5> <ul> <li>The Human Genome Project, initiated in the early 1990s, aimed to sequence the complete nucleotide content in human cells.</li> <li>The project was conducted by the National Human Genome Research Institute (NHGRI) of the National Institutes of Health, USA and the private company Celera Genomics, using different methods to sequence DNA.</li> <li>The publicly-funded initiative sequenced human DNA cloned into bacterial artificial chromosomes, while Celera Genomics used a whole-genome shotgun sequencing method on randomly cut human DNA.</li> <li>The first complete draft human genome was published in 2001, with the entire human genome released in 2003.</li> <li>The human genome sequence available today still has many gaps, with most of the sequenced part in the euchromatic region of the chromosomes and minimal representations from the heterochromatic regions.</li> <li>The heterochromatic areas, mainly consisting of repetitive elements, are located in the centromeric and telomeric regions of the chromosomes and are difficult to assay with the current sequencing technology.</li> </ul> <p>======</p> <h5 id="942-commonly-used-data-format">9.4.2 Commonly used data format</h5> <ul> <li>Bioinformatics often involves working with various data formats.</li> <li>A standard data format is generally used for the same analytes in the bioinformatics community, such as FASTQ format for most DNA sequencing data.</li> <li>FASTQ format is a text-based format that stores sequence information along with corresponding quality scores, with both encoded as single ASCII characters.</li> <li>FASTQ files typically have four lines per sequence, with the first line starting with a ‘$’ character followed by sequenced description, line 2 with raw sequence letters, line 3 with a ‘+’ character, and line 4 with quality values for the sequence in Line 2.</li> <li>FASTQ format uses FASTA format for sequence information, which is a text-based format representing sequence information in single-letter codes, with the first line starting with a ‘>’ symbol and holding a summary description of the sequence.</li> </ul> <p>======</p> <h5 id="943-genome-informatic-tools">9.4.3 Genome Informatic Tools</h5> <ul> <li>The text is about Quality Control (QC) for high-throughput sequence data.</li> <li>Two main tools mentioned are FastQC for QC and Trimmomatic for quality and adaptor trimming.</li> <li>The alignment-based workflow includes choosing a short or long read aligner, variant callers, post-processing, and interpreting filtered variants.</li> <li>Alignment of short sequence reads to a reference genome and visualization of three major types of variants (SNVs, InDels, CNVs) is demonstrated.</li> <li>Fig. 9.4A provides an example of visualizing SNVs, InDels, and CNVs on a read to reference alignment.</li> <li>No equations or formulae are provided in the text.</li> </ul> <p>======</p> <h5 id="introduction-to-bioinformatics">Introduction to Bioinformatics</h5> <p>• The text is about various types of tools used in different industries. • It explains the properties and uses of cutting tools, measuring tools, and clamping tools. • Cutting tools are further divided into single-point cutting tools and multipoint cutting tools. • Measuring tools are used to measure different parameters like length, angle, and temperature. • Clamping tools are used to hold the workpiece in place while machining operations are performed. • The text also discusses the importance of tool maintenance and safety measures while using these tools.</p> <p>======</p> <h5 id="944-downstream-analyses-interpreting-data-and-linking-to-function">9.4.4 Downstream analyses, interpreting data and linking to function</h5> <ul> <li>Genome annotation is the process of identifying and classifying all features in a genome, including coding genes, exon-intron junctions, repeat elements, non-coding elements, and pseudogenes. Repeat elements are classified into three major categories: transposons, satellites, and low complexity regions.</li> <li>Variant identification and classification involve reducing the number of variants called by variant callers through functional filters such as ones affecting protein sequence, splice variants, and stop codon read through. Validation of these variants in additional samples is required before interpreting their functional implications.</li> <li>Gene prediction involves identifying all coding elements in a genome using $a extit{b}$ initio, similarity-based, or integrated models. The similarity-based approach uses sequence similarity for identifying coding regions, while integrated models utilize both strategies to integrate gene prediction results.</li> <li>Gene ontology provides three vocabularies or ontologies that describe gene products regarding their associated biological processes, cellular components, and molecular functions.</li> <li>Transcript prediction is important to understand the identity of unannotated genes/transcripts. Tools like Cufflinks, cuffmerge, and cuffcompare help in the identification of transcripts.</li> <li>Phylogenetic analysis determines the evolutionary relationship of an organism with others through a cladogram or phylogenetic tree. Different methods for calculating evolutionary distances are used by tools for phylogenetic analysis like PHYLIP and PAUP.</li> </ul> <p>======</p> <h5 id="945-analysis-of-data-pertaining-to-human-genetic-diseases">9.4.5 Analysis of data pertaining to human genetic diseases</h5> <ul> <li>Genetic disorders can be caused by abnormalities in the genome, classified as monogenic (single gene) disorders, autosomal disorders, or sex chromosome-linked traits.</li> <li>Incomplete dominance or codominance exists when one allele is not entirely expressed over another, resulting in an intermediate phenotype, as seen in sickle cell anemia.</li> <li>Autosomal recessive variants in long stretches (1 MB or larger) are common in large consanguineous families, aiding in narrowing down disease-relevant variants.</li> <li>The analytical workflow for human genetics experiments involves sequencing reads, initial bioinformatics analysis, and filtering steps to distinguish disease-associated variants from common polymorphisms and sequencing errors.</li> <li>Exome sequencing, used extensively in the last 5 years, is cheaper and more straightforward than whole-genome sequencing for discovering human disease genes, though it focuses on protein coding regions and may miss non-coding variants.</li> <li>Whole genome sequencing offers a less biased and more comprehensive method to identify variants, including large structural variations, but may be expensive and require careful selection of healthy controls for rare genetic disorders.</li> </ul> <p>======</p> <h5 id="946-role-of-artificial-intelligence-ai-in-future">9.4.6 Role of Artificial Intelligence (AI) in future</h5> <ul> <li>Artificial Intelligence (AI) is making significant strides in various fields, including bioinformatics.</li> <li>AI-based tools are used in the USA for diagnosing eye diseases and have the potential to read X-rays and pathology slides, aiding radiologists and pathologists in correctly and quickly diagnosing human diseases.</li> <li>AI holds promise in helping farmers improve yield and decide on crops, and many existing bioinformatics tools use machine learning algorithms for calling genomic variants and assessing their significance levels.</li> <li>A successful intersection between AI and bioinformatics requires time and multidisciplinary efforts to build sufficient intelligence in the tools to interpret data and generate hypotheses.</li> <li>Collaboration among researchers in biology, computer science, bioinformatics, statistics, and artificial intelligence is essential for successful tools for biological data analysis in the future.</li> </ul> <p>======</p>