Bioinformatics

The rather vague term bioinformatics is a component of the new field of information science that uses statistical and other mathematical techniques to provide interpretations of experimental results obtained in the study of various problems. In bioinformatics the anphasis is on determining correlations in biology and medicine, e.g., in the prediction of disease probabilities. Bioinformatics is rapidly becoming a scientific disciple of its own, and it is not an easy one. 

A new subdiscipline of bioinformatics, comparative proteogenomics, is based on the possibility of mapping sets of sequenced peptides onto the positions in the genome that code for them, i.e., integration of data from MS-based proteomics with DNA sequence data sets. Databases and software to determine if the available data are over- or underrepresented include Gene Ontology, protein domains, and pathway databases. Another application (of many) is the combination of MS-based proteomic data with information from transcriptomics to study the processes and regulation of cellular functions. 

The following sections provide an overview on applications of expert system and related software in these fields. 

The analysis of chemicals by reference to a set of library mass spectra was facilitated by the establishment almost 40 years ago of databases such as the NIST/EPA/NIH reference library of electron impact mass spectra (http //www.nist.gov/srd/nistla.htm). Experimentally derived mass spectra are compared to spectra in the library, and the matches are graded by various algorithms. This comparison is valid because electron impact ionization requires that the sample be vaporized and thus isolated from its sampling history. 

For decades microorganisms have served as a source of proteins for structural and mechanistic protein biochemistry. Throughout the last decade, as new proteomics strategies were developed, these have usually been demonstrated 

The present chapter does not consider analysis of extracted protein biomarkers but rather focuses on strategies for rapid chemotaxonomic analysis of intact microorganisms with automated sample manipulation. Rapid means less than 5 minutes. Advantages of the application of bioinformatics and proteomics strategies for rapid identification of microorganisms include the following  

Proteomics algorithms have been developed to search databases of protein sequences for matches to short sequences determined experimentally from proteins or peptides in the laboratory,5 59 and for theoretical matches to 

These classical bioinformatics strategies are not all directly applicable to analysis of intact microorganisms, which each contain a mixture of protein biomarkers. Flowever, the principle of comparing experimental observations to protein/genome databases (bioinformatics) does hold the key to flexibility, 

Department of Statistics, North Carolina State University, Raleigh, North Carolina 27695 DAHLIA M. NIELSEN 

The field of bioinformatics has been defined in many ways and encompasses a large variety of areas of study. Within this scope, the common bond is genomic data, including its maintenance, analysis, and integration. In this chapter we retain this focus, providing an overview of various concepts of interest to researchers in the field of Toxicology. 

This chapter is informally divided into two parts. The first part is devoted to sequence comparisons and culminates in the database search tool BLAST. The second part considers the analysis of population-level data, providing sufficient background to understand how genes influencing traits such as response to toxicity may be located within the genome. 

Molecular and Biochemical Toxicology, Fourth Edition, edited by Robert C. Smart and Ernest Hodgson Copyright 2008 John Wiley Sons, Inc. 

While genes carry most of the information, it s the proteins that do the actual work. Humans appear to have only about 30,000 genes (wheat and barley have more). Over 40% of human proteins share similarity with proteins in fruit flies (Drosophila) or worms Caenorhabditis elegans). The science of proteomics seeks to identify the biochemical and cellular quantities, structures, and functions of all of these proteins, particularly in relation to their speciflc roles in disease. If it can be demonstrated that a particular protein is associated with a particular cellular function, then compounds which act on that protein may be useful in treating or cming a related disease. 

There is a constant demand for new algorithms, new and improved data visualization techniques, new computing platforms (such as grid computing), and new business models for the industry. 

Computational methods are essential in processing the vast data streams from sequencing efforts, stiTictural infomiation, protein analysis and differential gene expression. Computational approaches serve to integrate the research, development and discovery phases. An example of such efforts is the study of daig-likeness with respect to ADME parameters, in particular BBB penetration. 

The enormous task of determining the function and cooperation of proteins to create and maintain biological systems draws on various methods, e.g., patterns of co-occurrence identified from fused genes. 

Efficiency is achieved by a combination of experiments and computation, i.e. merging the results of expression profiling with DNA micro arrays with computation (co-occurrence, phylogenetic profiles and fused genes). 

Because there must be selective pressure for certain genes to be fused over the course of evolution, functional associations of proteins can be predicted. In total, 215 genes or proteins in the complete genomes of E. coli, Haemophilus influenza and Methanococcus jannaschii are involved in 64 unique fusion events. The approach is of general applicability and can be applied to genes of unknown function. 

Tlie availability of over 20 fully sequenced genomes has forced the development of new methods to find protem functions and interactions. Proteins were grouped by correlated evolution, correlated mRNA expression patterns and patterns of domain fusion to deter-mme functional relationships among the 6217 proteins of the yeast S. cerevisiae. Using these methods, over 93,000 pair-wise finks between functionally related yeast proteins were discovered. Links between characterized and uncharacterized proteins allow a general function to be assigned to more than half of the 2557 previously uncharacterized yeast proteins. 

---

Having settled on a definition of chemoinformatics, it is time for us to reflect on the distinction between chemoinformatics and bioinformatics. The objects of interest of bioinformatics are mainly genes and proteins. But genes, DNA and RNA, and proteins are chemical compounds They are objects of high interest in chemistry, Chemists have made substantial contributions to the elucidation of the structure and function of nucleic adds and proteins. The message is dear there is no clearcut distinction between bioinfonnatics and chemoinformatics I... 

Clearly, by tradition, chemoinformatics has largely dealt with small molecules, whereas bioinformatics has started to move from genes to proteins, compounds... 

We will make real progress in understanding the structure, the properties, and the function of proteins, DMA, and RNA only if bioinformatics and chemoinformatics work together ... 

In 1971 the Protein Data Bank - PDB [146] (see Section 5.8 for a complete story and description) - was established at Brookhaven National Laboratories - BNL -as an archive for biological macromolccular cr7stal structures. This database moved in 1998 to the Research Collaboratory for Structural Bioinformatics -RCSB. A key component in the creation of such a public archive of information was the development of a method for effreient and uniform capture and curation of the data [147], The result of the effort was the PDB file format [53], which evolved over time through several different and non-uniform versions. Nevertheless, the PDB file format has become the standard representation for exchanging inacromolecular information derived from X-ray diffraction and NMR studies, primarily for proteins and nucleic acids. In 1998 the database was moved to the Research Collaboratory for Structural Bioinformatics - RCSB. 

D images, and a variety of links to other resources bibliographic citations), the data entries are annotated by RCSB (Research CoUaboratory for Structural Bioinformatics) with additional information. 

The SWISS-PROT database [36] release 40.44 (February, 2003) contains over 120 000 sequences of proteins with more than 44 million amino adds abstracted from about 100 000 references. Besides sequence data, bibHographical references, and taxonomy data, there are highly valuable annotations of information (e.g., protein function), a minimal level of redundancy, and a high level of integration with other databases (EMBL, PDB, PIR, etc.). The database was initiated in 1987 by a partnership between the Department of Medicinal Biochemistry of the University of Geneva, Switzerland, and the EMBL. Now SWISS-PROT is driven as a joint project of the EMBL and the Swiss Institute of Bioinformatics (SIB). 

PDB Research Col-laboratory for Structural Bioinformat-ics (RCSB) macromole-cular structure data on proteins, nucleic acids, protein-nucleic acid complexes, and viruses nu- meric. biblio. -20000 records experi- ments Research Col-laboratory for Structural Bioinformatics online, CD-ROM periodi- cally WU7W.TCsh.0Tg/ pdh/... 

EMBL European Bioinformatics Institute nucleotide sequence database biblio., sub- stance, se- quence 20mio nucleotide seq., 28billion nucleotides journals, author submis- sions European Bioinformatics Institute free daily http //www.e- hi.ac.uk/embl/ index.html... 

We aim to show below how an explicit coding of the chemical structures of the starting materials and products of biochemical reactions and their reaction centers might allow us to achieve progress in our understanding of biochemical pathways. Furthermore, it will be shown how a bridge between chemoinformatics and bioinformatics can be built. 

Bioinformatics is a relatively new discipline that is concerned with the collection, organisatic and analysis of biological data. It is beyond our scope to provide a comprehensive overvie of this discipline a few textbooks and reviews that serve this purpose are now available (s the suggestions for further reading). However, we will discuss some of the main rnethoc that are particularly useful when trying to predict the three-dimensional structure and fum tion of a protein. To help with this. Appendix 10.1 contains a limited selection of some of tf common abbreviations and acronyms used in bioinformatics and Appendix 10.2 lists sorr of the most widely used databases and other resources. 

Appendix 10,1 Some Common Abbreviations and Acronyrris Used in Bioinformatics... 

A Zellner. An Introduction to Bayesian Inference in Econometrics. New York Wiley, 197L I Zhu, IS Em, CE Lawrence. Bayesian adaptive sequence alignment algorithms. Bioinformat-ics 14 25 -39, 1998. 

It is well known that the resources available on the Internet are in constant flux, with new sites appearing on a daily basis and established sites disappearing almost as frequently. This also holds true for the dedicated tools used in biochemical and biophysical studies. New tools are constantly becoming available, and established tools, obsolete. Such rapid change makes it difficult to stay current with the state-of-the-art technologies in the areas of bioinformatics and computational biochemistry and biophysics. 

Resource site for biotechnology—Molecular biology, bioinformatics, biophysics, and biochemistry—A well-organized web site http //WWW.ahpcc. unm. edu/ aroberts/... 

---