Content uploaded by Ardeshir Bayat
Author content
All content in this area was uploaded by Ardeshir Bayat
Content may be subject to copyright.
Clinical review
Science, medicine, and the future
Bioinformatics
Ardeshir Bayat
An unprecedented wealth of biological data has been
generated by the human genome project and sequenc-
ing projects in other organisms. The huge demand for
analysis and interpretation of these data is being man-
aged by the evolving science of bioinformatics.
Bioinformatics is defined as the application of tools of
computation and analysis to the capture and interpret-
ation of biological data. It is an interdisciplinary field,
which harnesses computer science, mathematics, phys-
ics, and biology (fig 1). Bioinformatics is essential for
management of data in modern biology and medicine.
This paper describes the main tools of the bioinforma-
tician and discusses how they are being used to
interpret biological data and to further understanding
of disease. The potential clinical applications of these
data in drug discovery and development are also
discussed.
Methods
This article is based on personal experience in
bioinformatics and on selected articles in recent issues
of Nature Genetics,Nature Genetics Reviews,Nature Medi-
cine, and Science. Key terms including bioinformatics,
comparative and functional genomics, proteomics,
microarray, disease, and medicine were used to search
for relevant articles in the peer reviewed scientific
literature.
Bioinformatics and its impact on
genomics
Last year it was announced that the entire human
genome had been mapped as a result of the efforts of
the worldwide human genome project and a private
genomic company.12However, in recent years, the sci-
entific world has witnessed the completion of whole
genome sequences of many other organisms. The
analysis of the emerging genomic sequence data and
the human genome project is a landmark achievement
for bioinformatics.
A novel strategy for random sequencing of the
whole genome (the so called “shot gun” technique) was
used to sequence the genome of Haemophilus influenzae
in 1995.3This was the very first complete genome of
any free living organism to be sequenced. Other bacte-
rial genomes, such as those of Mycoplasma genitalium
and Mycobacterium tuberculosis, were sequenced soon
after,45and the sequence of the plague bacterium Ye r s -
inia pestis was recently completed.6The sequence and
annotation of the first eukaryotic genome, that of Sac-
charomyces cerevisiae (a yeast),7was followed by those of
other eukaryotic species such as Caenorhabtidis elegans
(a worm),8Drosophila melanogaster (fruit fly),9and Arab-
dopsis thaliana (mustard weed)10 (see fig A on bmj.com).
Sequencing of several other species, including
An additional figure
appears on
bmj.com
Biology Medicine
Maths/physics Computer science
Bioinformatics
Fig 1 Interaction of disciplines that have contributed to the
formation of bioinformatics
Summary points
Bioinformatics is the application of tools of
computation and analysis to the capture and
interpretation of biological data
Bioinformatics is essential for management of
data in modern biology and medicine
The bioinformatics toolbox includes computer
software programs such as BLAST and Ensembl,
which depend on the availability of the internet
Analysis of genome sequence data, particularly
the analysis of the human genome project, is one
of the main achievements of bioinformatics to
date
Prospects in the field of bioinformatics include its
future contribution to functional understanding
of the human genome, leading to enhanced
discovery of drug targets and individualised
therapy
Bioinformatics,
a new
interdisciplinary
science, is
essential to
managing,
understanding,
and harnessing
clinical benefit
from new
genetic data
Centre for
Integrated Genomic
Medical Research,
University of
Manchester,
Manchester
M13 9PT
Ardeshir Bayat
MRC fellow
Correspondence to:
ardeshir.bayat@
man.ac.uk
BMJ 2002;324:1018–22
1018 BMJ VOLUME 324 27 APRIL 2002 bmj.com
zebrafish, pufferfish, mouse, rat, and non-human
primates, are either under way or nearing completion
by both private and public sequencing initiatives.11 The
knowledge obtained from these sequence data will
have considerable implications for our understanding
of biology and medicine. As a result of comparative
genomic and proteomic research, we will soon be able
to not only locate each human gene but also fully
understand its function.12
Bioinformatic tools
The main tools of a bioinformatician are computer
software programs and the internet. A fundamental
activity is sequence analysis of DNA and proteins using
various programs and databases available on the world
wide web. Anyone, from clinicians to molecular
biologists, with access to the internet and relevant web-
sites can now freely discover the composition of
biological molecules such as nucleic acids and proteins
by using basic bioinformatic tools. This does not imply
that handling and analysis of raw genomic data can
easily be carried out by all. Bioinformatics is an evolv-
ing discipline, and expert bioinformaticians now use
complex software programs for retrieving, sorting out,
analysing, predicting, and storing DNA and protein
sequence data.
Large commercial enterprises such as pharmaceu-
tical companies employ bioinformaticians to perform
and maintain the large scale and complicated bioinfor-
matic needs of these industries. With an ever-
increasing need for constant input from bioinformatic
experts, most biomedical laboratories may soon have
their own in-house bioinformatician. The individual
researcher, beyond a basic acquisition and analysis of
simple data, would certainly need external bioinfor-
matic advice for any complex analysis.
The growth of bioinformatics has been a global
venture, creating computer networks that have allowed
easy access to biological data and enabled the develop-
ment of software programs for effortless analysis.
Multiple international projects aimed at providing
gene and protein databases are available freely to the
whole scientific community via the internet.
Bioinformatic analysis
The escalating amount of data from the genome
projects has necessitated computer databases that fea-
ture rapid assimilation, usable formats and algorithm
software programs for efficient management of
biological data.13 Because of the diverse nature of
emerging data, no single comprehensive database
exists for accessing all this information. However, a
growing number of databases that contain helpful
information for clinicians and researchers are avail-
able. Information provided by most of these databases
is free of charge to academics, although some sites
require subscription and industrial users pay a licence
fee for particular sites. Examples range from sites pro-
viding comprehensive descriptions of clinical disor-
ders, listing disease susceptibility genetic mutations
and polymorphisms, to those enabling a search for dis-
ease genes given a DNA sequence (box).
These databases include both “public” repositories
of gene data as well as those developed by private com-
panies. The easiest way to identify databases is by
Useful bioinformatic websites (available freely
on the internet)
•National Center for Biotechnology Information
(www.ncbi.nlm.nih.gov)
—
maintains bioinformatic tools
and databases
•National Center for Genome Resources
(www.ncgr.org/)
—
links scientists to bioinformatics
solutions by collaborations, data, and software
development
•Genbank (www.ncbi.nlm.nih.gov/Genbank)
—
stores
and archives DNA sequences from both large scale
genome projects and individual laboratories
•Unigene (www.ncbi.nlm.nih.gov/UniGene)
—
gene
sequence collection containing data on map location
of genes in chromosomes
•European Bioinformatic Institute
(www.ebi.ac.uk)
—
centre for research and services in
bioinformatics; manages databases of biological data
•Ensembl (www.ensembl.org)
—
automatic annotation
database on genomes
•BioInform (www.bioinform.com)
—
global
bioinformatics news service
•SWISS-PROT (www.expasy.org/sprot/)
—
important
protein database with sequence data from all
organisms, which has a high level of annotation
(includes function, structure, and variations) and is
minimally redundant (few duplicate copies)
•International Society for Computational Biology
(www.iscb.org/)
—
aims to advance scientific
understanding of living systems through computation;
has useful bioinformatic links
Fig 2 Ensembl website: a genomic data search facility freely available on the internet.
Ensembl is a joint project between the European Bioinformatic Institute and the Sanger
Centre, which is capable of automatically tracking the sequenced pieces of the human
genome and assembling and analysing them to identify genes and other features of interest
to biomedical researchers
Clinical review
1019BMJ VOLUME 324 27 APRIL 2002 bmj.com
searching for bioinformatic tools and databases in any
one of the commonly used search engines. Another
way to identify bioinformatic sources is through
database links and searchable indexes provided by one
of the major public databases. For example, the
National Center for Biotechnology Information
(www.ncbi.nlm.nih.gov) provides the Entrez browser,
which is an integrated database retrieval system that
allows integration of DNA and protein sequence data-
bases. The European Bioinformatic Institute archives
gene and protein data from genome studies of all
organisms, whereas Ensembl produces and maintains
automatic annotation on eukaryotic genomes (fig 2).
The quality and reliability of databases vary; certainly
some of the better known and more established ones,
such as those above, are superior to others.
One of the simplest and better known search tools
is called BLAST (basic local alignment search tool, at
www.ncbi.nlm.nih.gov/BLAST/). This algorithm soft-
ware is capable of searching databases for genes with
similar nucleotide structure (fig 3) and allows compari-
son of an unknown DNA or amino acid sequence with
hundreds or thousands of sequences from human or
other organisms until a match is found. Databases of
known sequences are thus used to identify similar
sequences, which may be homologues of the query
sequence. Homology implies that sequences may be
related by divergence from a common ancestor or
share common functional aspects. When a database is
searched with a newly determined sequence (the query
sequence), local alignment occurs between the query
sequence and any similar sequence in the database.
The result of the search is sorted in order of priority on
the basis of maximum similarity.The sequence with the
highest score in the database of known genes is the
homologue. If homologues or related molecules exist
for a query sequence, then a newly discovered protein
may be modelled and the gene product may be
predicted without the need for further laboratory
experiments.
Functional genomics
Since the completion of the first draft of the human
genome,12 the emphasis has been changing from
genes themselves to gene products. Functional genom-
ics assigns functional relevance to genomic infor-
mation. It is the study of genes, their resulting proteins,
and the role played by the proteins.
Analysis and interpretation of biological data con-
siders information not only at the level of the genome
but at the level of the proteome and the transcriptome
(fig 4). Proteomics is the analysis of the total amount of
proteins (proteome) expressed by a cell, and transcrip-
tomics refers to the analysis of the messenger RNA
transcripts produced by a cell (transcriptome). DNA
microarray technology determines the expression level
of genes and includes genotyping and DNA sequenc-
ing. Gene expression arrays allow simultaneous analy-
sis of the messenger RNA expression levels of
thousands of genes in benign and malignant tumours,
such as keloid and melanoma. Expression profiles clas-
sify tumours and provide potential therapeutic
targets.14
Bioinformatic protein research draws on annotated
protein and two dimensional electrophoresis data-
bases. After separation, identification, and characterisa-
tion of a protein, the next challenge in bioinformatics is
the prediction of its structure. Structural biologists also
Fig 3 Web page illustrating freely available BLAST services run by
the National Center for Biotechnology Information. BLAST (basic local
alignment search tool) is a set of similarity search programs
designed to explore all of the available DNA sequence databases
Genomics
...agcttgatattatacgcgcggca
Transcriptomics
Proteomics
DNA
I
II
IV
III
RNA
Protein
Makes
Makes
C
o
m
p
l
e
x
i
t
y
Fig 4 Schematic diagram representing complexity of genomic data processing. Analysis and
interpretation of biological data considers information at every level from the genome (total
genetic content) to the proteome (total protein content) and transcriptome (total messenger
RNA content) of the cell. The images numbered I-IV to the right of the diagram represent
relevant examples of DNA (image I is base pair nucleotides); RNA (image II is a microarray
showing levels of gene expression); and protein (image III is a structure of a single protein;
image IV is a two dimensional gel electrophoresis showing separation of all proteins of a
cell—each spot corresponds to a different protein chain)
Clinical review
1020 BMJ VOLUME 324 27 APRIL 2002 bmj.com
use bioinformatics to handle the vast and complex data
from xray crystallography, nuclear magnetic reso-
nance, and electron microscopy investigations to create
three dimensional models of molecules.15
Other applications of bioinformatics
Apart from analysis of genome sequence data,
bioinformatics is now being used for a vast array of
other important tasks, including analysis of gene varia-
tion and expression, analysis and prediction of gene
and protein structure and function, prediction and
detection of gene regulation networks, simulation envi-
ronments for whole cell modelling, complex modelling
of gene regulatory dynamics and networks, and
presentation and analysis of molecular pathways in
order to understand gene-disease interactions.16
Although on a smaller scale, simpler bioinformatic
tasks valuable to the clinical researcher can vary from
designing primers (short oligonucleotide sequences
needed for DNA amplification in polymerase chain
reaction experiments) to predicting the function of
gene products.
Clinical application of bioinformatics
The clinical applications of bioinformatics can be
viewed in the immediate, short, and long term. The
human genome project plans to complete the human
sequence by 2003, producing a database of all the vari-
ations in sequence that distinguish us all. The project
could have considerable impact on people living in
2020
—
for example, a complete list of human gene
products may provide new drugs and gene therapy for
single gene diseases may become routine
(www.ornl.gov/hgmis/medicine/tnty.html).
Basic bioinformatic tools are already accessed in
certain clinical situations to aid in diagnosis and treat-
ment plans. For example, PubMed (www.nlm.nih.gov)
is accessed freely for biomedical journals cited in
Medline, and OMIM (Online Mendelian Inheritance in
Man at www3.ncbi.nlm.nih.gov/Omim/), a search tool
for human genes and genetic disorders, is used by cli-
nicians to obtain information on genetic disorders in
the clinic or hospital setting. An example of the appli-
cation of bioinformatics in new therapeutic advances is
the development of novel designer targeted drugs such
as imatinib mesylate (Gleevec), which interferes with
the abnormal protein made in chronic myeloid
leukaemia.17 (Imatinib mesylate was synthesised at
Novartis Pharmaceuticals by identifying a lead in a
high throughput screen for tyrosine kinase inhibitors
and optimising its activity for specific kinases.) The
ability to identify and target specific genetic markers by
using bioinformatic tools facilitated the discovery of
this drug.
In the short term, as a result of the emerging bioin-
formatic analysis of the human genome project, more
disease genes will be identified and new drug targets
will be simultaneously discovered. Bioinformatics will
serve to identify susceptibility genes and illuminate the
pathogenic pathways involved in illness, and will there-
fore provide an opportunity for development of
targeted therapy. Recently, potential targets in cancers
were identified from gene expression profiles.18
In the longer term, integrative bioinformatic analy-
sis of genomic, pathological, and clinical data in clinical
trials will reveal potential adverse drug reactions in
individuals by use of simple genetic tests. Ultimately,
pharmacogenomics (using genetic information to
individualise drug treatment) is likely to bring about a
new age of personalised medicine; patients will carry
gene cards with their own unique genetic profile for
certain drugs aimed at individualised therapy and tar-
geted medicine free from side effects.
Future directions
The practice of studying genetic disorders is changing
from investigation of single genes in isolation to
discovering cellular networks of genes, understanding
their complex interactions, and identifying their role in
disease.19 As a result of this, a whole new age of
individually tailored medicine will emerge. Bioinfor-
matics will guide and help molecular biologists and
clinical researchers to capitalise on the advantages
brought by computational biology.20 The clinical
Additional educational resources
Journals
•Specific bioinformatic journals exist (for example,
www.bioinformatics.oupjournals.org), but papers from
every area of science and medicine involving
bioinformatic analysis are published in any biomedical
journal. Examples include:
•The human genome (special issue). Nature
2001;409:813-933.
•The human genome (special issue). Science
2001;5507:1145-434.
•The human genome (special issue). JAMA
2001;286:2211-333.
•The human genome (special issue). Scientific
American 2000;283:38-57.
•Luscombe NM, Greenbaum D,Gerstein M. What is
bioinformatics? Method Inform Med 2001;40:346-58.
•Online Lectures on Bioinformatics
(www.lectures.molgen.mpg.de/)
Books
•Mount DW. Bioinformatics: sequence and genome
analysis. Cold Spring Harbor Laboratory Press, 2001.
•Baxevanis AD, Ouellette BFF. Bioinformatics: a
practical guide to the analysis of genes and proteins. 2nd ed.
John Wiley and Sons, 2001.
•Lengauer T (Ed). Bioinformatics. Wiley-VCH Series,
2001. (Methods and principles in medicinal chemistry
series.)
•Higgins D, Taylor W. Bioinformatics. Oxford
University Press, 2000. (Practical approach series.)
•Baldi P, Brunak S. Bioinformatics. 2nd ed. MIT Press,
2001. (Adaptive computation and machine learning
series.)
BMJ archive
•Aitman TJ. DNA microarrays in medical practice.
BMJ 2001;323:611-5.
•Mathew CG. Postgenomic technologies: hunting the
genes for common disorders. BMJ 2001;322:1031-4.
•Stewart A, Haites N, Rose P. Online medical genetics
resources: a UK perspective. BMJ 2001;322:1037-9.
•Savill J. Molecular genetic approaches to
understanding disease. BMJ 1997;314:126-9.
Clinical review
1021BMJ VOLUME 324 27 APRIL 2002 bmj.com
research teams that will be most successful in the com-
ing decades will be those that can switch effortlessly
between the laboratory bench, clinical practice, and the
use of these sophisticated computational tools.
I thank Tessa Richards, Dipak Roy, and Professor Bill Ollier for
advice on the preparation of this manuscript and Andy Brass for
providing me with some of the diagrams.
Funding: Medical Research Council.
Competing interests: None declared.
1 Inter national Human Genome Sequencing Consortium. Initial sequenc-
ing and analysis of the human genome. Nature 2001;409:860-921.
2 Venter JC, Adams MD, Myers EW, Li PW,Mural RJ, Sutton GG, et al. The
sequence of the human genome. Science 2001;291:1304-51.
3 Fleischmann RD, Adams MD, White O, Clayton RA, Kirkness EF, Kerlav-
age AR, et al. Whole-genome random sequencing and assembly of Hae-
mophilus influenzae Rd. Science 1995;269:496-512.
4 Fraser CM, Gocayne JD, White O, Adams MD, Clayton RA, Fleischmann
RD, et al. The minimal gene complement of Mycoplasma genitalium.
Science 1995;270:397-403.
5 Cole ST,Brosch R, Parkhill J, Gar nier T,Churcher C, Har ris D, et al. Deci-
phering the biology of Mycobacterium tuberculosis from the complete
genome sequence. Nature 1998;393:537-44.
6 Parkhill J, Wren BW,Thomson NR, Titball RW, Holden MT, Prentice MB,
et al. Genome sequence of Yersinia pestis, the causative agent of plague.
Nature 2001;413:523-27.
7 Goffeau A, Barrell BG, Bussey H, Davis RW, Dujon B, Feldmann H, et al.
Life with 6000 genes. Science 1996;274:546.
8 The C. elegans Sequencing Consortium. Genome sequence of the nema-
tode C. elegans: a platform for investigating biology. Science
1998;282:2012-8.
9 Myers EW, Sutton GG, Delcher AL, Dew IM, Fasulo DP, Flanigan MJ, et al.
A whole-genome assembly of Drosophila. Science 2000;287:2196-204.
10 Arabidopsis Genomics Initiative. Analysis of the genome sequence of the
flowering plant Arabidopsis thaliana. Nature 2000;408:796-815.
11 Stein L. Genome annotation: from sequence to biology. Nat Rev Genet
2001;2:493-503.
12 Subramanian G, Adams MD, Venter JC, Broder S. Implications of the
human genome for understanding human biology and medicine. JAMA
2001;286:2296-306.
13 Benton D. Bioinformatics
—
principles and potential of a new multidisci-
plinary tool. Trends Biotech 1996;14:261-312.
14 Maggio ET, Ramnarayan K. Recent developments in computational pro-
teomics. Trends Biotech 2001;19:266-72.
15 Burley SK, Almo SC, Bonanno JB, Capel M, Chance MR, Gaasterland T,
et al. Structural genomics: beyond the human genome project. Nat Genet
1999;23:151-7.
16 Tsoka S, Ouzounis CA. Recent developments and future directions in
computational genomics. FEBS Lett 2000;480:42-8.
17 Druker BJ, Sawyers CL, Kantarjian H, Resta DJ, Reese SF, Ford JM, et al.
Activity of a specific inhibitor of the BCR-ABL tyrosine kinase in the blast
crisis of chronic myeloid leukemia and acute lymphoblastic leukemia
with the Philadelphia chromosome. N Engl J Med 2001;344:1038-42.
18 Graeber TG, Eisenberg D. Bioinformatic identification of potential auto-
crine signaling loops in cancers from gene expression profiles. Nat Genet
2001;29:295-300.
19 Debouk C, Metcalf B. The impact of genomics on drug discovery. Annu
Rev Pharmacol Toxicol 2000;40:193-208.
20 Butler D. Are you ready for the revolution? Nature 2001;409:758-60.
When I use a word
Meta-
Mr John Gleave, a neurosurgeon,has written to ask me
the origin of the meta- in meta-analysis. The answer
comes from Aristotle.
The Greek preposition ìåôá´ (meta) had several
meanings, depending on whether it governed the
accusative, genitive,or dative case. With the accusative
it could mean coming into or among, in pursuit of, or
coming after in place or time; with the genitive it could
mean in the midst of, between, or in common with;
and with the dative it could mean in the company of or
over and above. It was also used as a prefix to express
such notions as sharing, being in the midst of,
succession, pursuit, reversal, and (most commonly)
change. Examples of the last include metabolism,
metamorphosis, and metaplasia.
In scientific English words its uses include
“consequent upon” (as in the obsolete terms
meta-arthritic, metapneumonic), “behind” or “beyond”
in an anatomical sense (metabranchial, metacarpal,
metaphysis), “coming later” (metaphase, which comes
after prophase), or “changing” (metachromasia, a
property of materials that stain a different colour from
the stain used). In geology meta- is used to distinguish
various types of metamorphic processes. And chemists
use meta- to differentiate certain metameric chemical
compounds (such as metacresol, paracresol,
orthocresol).
And so to Aristotle. Some 250 years after his death,
Aristotle’s manuscripts came into the hands of
Andronicus of Rhodes, who edited them. Andronicus
called one set of papers The Physics (ôá`*ïõóéêá´), dealing
as they did with natural science. Then he published a
set of papers that he called The Metaphysics (ôá`ìåôá`ôá`
*ïõóéêá´), simply because it came after The Physics.
However, because The Metaphysics dealt with what
Aristotle called “primary philosophy,” or ontology,
metaphysics came to be misunderstood as “the science
of that which transcends the physical.”
As a result, the prefix meta- was then used to
designate any higher science (actual or hypothetical)
that dealt with more fundamental problems than the
original science itself. This use first appeared in the
early 17th century (John Donne, for example, writes
about metatheology) but did not become really
popular until the middle of the 19th century. Examples
include metaethics (the study of the foundations of
ethics, especially the nature of ethical statements) and
metahistory (an inquiry into the principles that govern
historical events).
Then, from about 1940, it became commonplace to
prefix meta- to designate concern with basic principles.
A metacriterion is a criterion that defines criteria. A
metatheorem is a theorem about theorems. A
metalanguage is a language that supplies terms for
analysing a language; a metametalanguage does the
same for a metalanguage. And Jean Tinguely described
his machine-like sculptures as “metamechanical.”(But
a metaphysician is not a doctor’s doctor.)
In these poststructuralist times we recognise many
metaforms. Mantissa, a medical novel by John Fowles, is
metafiction; Francois Truffaut’s film La Nuit Amercaine
is metacinema; several paintings by Magritte, notably
La Condition Humaine, are meta-art; and John Cage’s
piano piece 4’33’’ is metamusic.
So meta-analysis is an analysis of analyses, in which
sets of previously published (or unpublished) data are
themselves subjected as a whole to further analysis. In
this statistical sense it was first used in the 1970s by GV
Glass (Educ Res 1976;3(Nov):2). As he wrote, “The term
is a bit grand, but it is precise and apt.” Incidentally,
meta-analysis should not be confused with metanalysis,
which is the process whereby, for example,“a nadder”
becomes “an adder” (see BMJ 1999;318:1758 and
2000;321:953).
I trust that this cures Mr Gleave’s metagrobolism.
Jeff Aronson clinical pharmacologist, Oxford
We welcome articles up to 600 words on topics such as
A memorable patient, A paper that changed my practice,My
most unfortunate mistake, or any other piece conveying
instruction, pathos, or humour.If possible the article
should be supplied on a disk. Permission is needed
from the patient or a relative if an identifiable patient is
referred to.
Clinical review
1022 BMJ VOLUME 324 27 APRIL 2002 bmj.com
