1
Introduction to Bioinformatics

1.1 Introduction

Interesting research fields emerge through the collaboration of researchers from different, sometimes distant, disciplines. Examples include biochemistry, biophysics, quantum information science, systems engineering, mechatronics, business information systems, management information systems, geophysics, biomedical engineering, cybernetics, art history, media technology and others. This marriage between disciplines yields findings which blend the views of different areas over the same subject or set of data.

The stimuli leading to such collaborations are numerous. For example, one discipline may develop tools that generate types of data that require another discipline to analyse. In other cases, one field scratches a layer of unknowns to discover that significant parts of its scope are actually based on the principles of another field, such as the low-level biological studies of the chemical interactions in the cells, which delivered biochemistry as an interdisciplinary field. Other interdisciplinary fields emerged because of their complementary involvement in building different parts of the same target system or in understanding different sides of the same research question; for example, mechatronics engineering aims at building systems which have both mechanical and electronic parts, such as all modern automobiles. Interdisciplinary areas like business information systems and management information systems have emerged due to the high demand for information systems which target business and management aspects; although generic information systems would meet many of those requirements, a customised field focusing on such applications is indeed more efficient given such high demand.

The interdisciplinary field of this book's focus is bioinformatics. The motive behind this field's emergence is the increasingly expanding generation of massive raw biological data following the developments in high-throughput techniques in the last couple of decades. The scale of this high-throughput data is orders of magnitude higher than what can be efficiently analysed in a manual fashion. Consequently, information engineers were recruited in order to contribute to data analysis by employing their computational methods. Cycles of computational analysis, sharing of results, interdisciplinary discussions and abstractions have led, and are still leading, to many key discoveries in biology and medicine. This success has attracted many information engineers towards biology and many biologists towards information engineering to meet in a potentially rich intersection area, which itself has grown in size to establish the field of bioinformatics.

1.2 The "Omics" Era

A new suffix has been introduced to the English language in this era of high-throughput data expansion; that is "-omics", and its relatives "-ome" and "-omic". This started in the 1930s when the entire set of genes carried by a chromosome was called the genome, blending the words "gene" and "chromosome" (OED, 2014). Consequently, the analysis of the entire genome was called genomics, and many known research journals carried the term "genome" or "genomics" in their titles such as Genomics, Genome Research, Genome Biology, BMC Genomics, Genome Medicine, the Journal of Genetics and Genomics (JGG), and others.

The -ome suffix was not kept exclusive for the genome; it has been rather generalised to indicate the complete set of some type of molecule or object. The proteome is the complete set of proteins in a cell, tissue or organism. Similarly are the transcriptome, metabolome, glycome and lipidome for the complete sets of transcripts, metabolites, glycans (carbohydrates) and lipids. In a respective order, large-scale studies of those complete sets are known as proteomics, transcriptomics, metabolomics, glycomics and lipidomics. The -ome suffix was further generalised to include the complete sets of objects other than basic molecules. For example, the microbiome is the complete set of microorganisms (e.g. bacteria, microscopic fungi, etc.) in a given environment such as a building, a sample of soil or the human gut (Kembel et al., 2014). More omic fields have also emerged such as agrigenomics (the application of genomics in agriculture), pharmacogenomics and pharmacoproteomics (the application of genomics and proteomics to pharmacology), and others.

All of those biological fields of omics involve high-throughput datasets which are subject to information engineering involvements, and therefore reside at the core focus of bioinformatic research. An even higher level of omics analysis involves integrative analysis of many types of omic datasets. OMICS: a Journal of Integrative Biology is a journal which targets research studies that consider such collective analysis at different levels from single cells to societies.

More types of high-throughput omic datasets are expected to emerge. The role of bioinformatics as an interdisciplinary field will be more important. This is not only because each of those omic datasets is massive in size when considered individually; it is also because of the size of information hidden in the relations between those generally heterogeneous datasets, which requires more sophisticated computational methods to analyse.

1.3 The Scope of Bioinformatics

The scope of bioinformatics includes the development of methods, techniques and tools which target storage, retrieval, organisation, analysis and presentation of high-throughput biological data.

1.3.1 Areas of Molecular Biology Subject to Bioinformatics Analysis

In a very general statement, each part of molecular biology which produces high-throughput data is subject to bioinformatics analysis. On the other hand, low-throughput data which can be manually analysed do not represent subjects for bioinformatics. The omics fields described in the previous section are indeed included in bioinformatics analysis. This includes aspects of DNA, RNA and protein sequence analysis, gene and protein expression, genetics of diseases including cancers and special phenotypes, analysis of gene regulation, chemical interaction regulation, enzymatic regulation, other types of regulation, analysis of flowing signals in cells, networks of genetic, protein and other molecular interactions, comparable analysis of the diversity of genomes between individuals or organisms in an environment or across different environments, and others.

1.3.2 Data Storage, Retrieval and Organisation

The human genome is a linear thread of more than three billion base-pairs (letters). In 2012, and after more than 4 years after its starting point, the 1000 Genomes Project Consortium announced the completion of sequencing of the complete genomes of 1092 individuals from fourteen different populations (The 1000 Genomes Project Consortium, 2014). Moreover, the genomes of thousands of organisms, other than humans, have been sequenced and stored during the last two to three decades. As for gene expression data, tens of thousands of massive microarray datasets have been generated in the last two decades. Add to that the increasing amounts of data generated for protein expression, DNA binding and other types of high-throughput data. Data generation has not stopped and is expected to increase rapidly due to the massive advances in technologies and cost reduction. Therefore, it is crucial to store such amounts of datasets in an efficient manner which allows for quick and efficient access by large numbers of researchers from different parts of the world simultaneously.

Given the current trend, which is to offer most of the generated high-throughput datasets for public use in centralised databases, it becomes essential to standardise the way in which data are organised, annotated and labelled. This enhances information exchange and mutual understanding between different research groups in the world.

Taken together, the scope of bioinformatics indeed includes designing and implementing appropriate databases for high-throughput biological data storage, building means of data access to those databases such as web services and network applications, organising different levels of data pieces by standard formats and annotations, and, undoubtedly, maintaining and enhancing the availability and the scalability of these data repositories.

1.3.3 Data Analysis

Elaine Mardis, the Professor of Genetics in the Genome Institute at Washington University, and a collaborator in the 1000 Genomes Project, titled her "musing" published in Genome Medicine in 2010 as "the $1,000 genome, the $100,000 analysis?" (Mardis, 2010). Mardis discussed the tremendous drop in the cost of sequencing the complete genome of an individual human from hundreds of millions of dollars to a few thousands, and that it is expected to reach the line of $1,000. She mused, based on many facts and observations, that the cost of data analysis, which does not seem to be dropping, will constitute the major part of the total cost, rather than the cost of data generation.

A small proportion of human genes, out of 20 000-25 000, has been well described and understood, while many gaps in our understanding of the vast majority of them do still exist. Identifying the sequence of a gene from a thousand individuals and measuring...