Bioinformatics FAQ

From Bioinformatics.Org Wiki

Revision as of 19:58, 13 July 2007 by Sonny (Talk | contribs)
Jump to: navigation, search

We're in the process of moving the subsections to separate articles. Please pardon the mess.



What is bioinformatics?

What are the origins of bioinformatics?

What are the most common bioinformatics programs?

What are the most common bioinformatics technologies?

How are data analyzed in bioinformatics?

Fields Related to Bioinformatics

What is biophysics?

What is computational biology?

What is medical informatics?

What is cheminformatics?

What is genomics?

What is mathematical biology?

What is proteomics?

What is pharmacogenomics?

What is pharmacogenetics?

Books: Can you recommend any bioinformatics books?

It's notoriously difficult to find any books on bioinformatics itself that cater well for all of those coming from computing, from mathematics and from biology backgrounds. The few textbooks available in the field tend to be eyewateringly expensive as well. I've divided suggested reading into [#generalBooks books of general interest], [#computerScientistsBooks those] best suited to people coming from a computational/mathematical background and [#biologistsBooks books for biologists interested in bioinformatics]. After my suggestions are some links to other lists of bioinformatics books.

General introductions

Many people are curious about the Human Genome (Project). The completion of the first draft probably represents bioinformatics' coming of age as a discipline. The first couple of books are aimed at the intelligent layperson.

A gossipy and insightful account of the race to sequence the genome can be found in "The Sequence" by Kevin Davies [Weidenfeld; ISBN 0297646982]. Matt Ridley's "Genome" [Fourth Estate; ISBN 185702835X] is both an interesting layperson's introduction to the issues raised by the bioinformatic revolution and an overview of its biology and enormous scope. If I remember rightly, Ridley's book received a slightly snooty review from Walter Bodmer. This is understandable, since his and Robin McKie's excellent "pre-genomic" guide to the Human Genome Mapping Project, "The Book of Life" [Oxford Paperbacks; ISBN 0195114876] was undeservedly in a remainders bin when I bought my copy a couple of years ago.

If you are a non-biological scientist (or a non-scientist) and are hooked by these, why not go back to the "real beginning" of the race and read James Watson's entertaining and indiscreet memoir of his and Francis Crick's determination of the structure of DNA, "The Double Helix" [Penguin; ISBN 0140268774]---now updated with an introduction by media don Steve Jones.

Nigel Barber at Peterborough Regional College in the UK recommends Gary Zweiger's "Transducing the Genome" [McGraw-Hill Professional Publishing: ISBN 0071369805]. The summary at Amazon makes it sound a tad pretentious, but all the reviews seem pretty positive so it might be worth a read.

If you are a quantitative scientist and would like a deeper knowledge of contemporary (molecular) biology, but you want to acquire it as painlessly as possible you could try the following:

Computational/Mathematical aspects

If you are a hardcore maths/computing person Michael Waterman's "Introduction to Computational Biology" [Chapman & Hall/CRC Statistics and Mathematics; ISBN 0412993910] and Pavel Pevzner's "Computational Molecular Biology - An Algorithmic Approach" [The MIT Press (A Bradford Book); ISBN 0262161974] will give you all the discrete maths you can shake a stick at, but perfunctory introductions to the biology.

Bioinformatics.Org's very own Jeff Bizzaro recommends Dan Gusfield's "Algorithms on Strings, Trees and Sequences" [Cambridge, 1997 ISBN 0-52158-519-8], Richard Durbin, S. Eddy, A. Krogh, G. Mitchison "Biological Sequence Analysis: Probabilistic Models of Proteins and Nucleic Acids" [Cambridge, 1997 ISBN 0-52162-971-3] (which I think is one of the clearest and most comprehensive guides to alignment algorithms) and---for that full "computers-to-biology conversion"--- Geoffrey M. Cooper "The Cell: A Molecular Approach" [ASM Press, 1996 ISBN 0-87893-119-8]. Jeff Ames writes that a second edition of this book is now available [Sinauer Associates, Incorporated, 2000 ISBN 0-87893-106-6] and that this version---if you can find it in the shops---comes with a CD.

Applying bioinformatics to biological research

One outstanding general text for the biologist is David W. Mount's "Bioinformatics" [Cold Spring Harbor Press; ISBN 0879696087]. It's not cheap, but it's the best I've seen if you are studying bioinformatics itself.

Bioinformatics has been dismissed by some as "the science of BLAST searches". The best collection of advice so far on doing BLAST searches is O'Reilly's BLAST book by Ian Korf, Mark Yandell and Joseph Bedell [O'Reilly ISBN 0-596-00299-8]. I reviewed it enthusiastically, but not uncritically, for the UK UNIX Users' Group magazine. I'd go as far as to say that all biologists thinking of using BLAST in their research should read the relevant sections before they even go near a computer.

If you wish to use general bioinformatics tools, especially if you are a little wary of computers, my new "best" book is "Bioinformatics for Dummies" [John Wiley and Sons ISBN 0764516965]. It is (obviously) aimed at people who are beginners, who are happier using the Web rather than typing commands, and who are more interested in learning than in impressing people---the writing is friendly clear and unpretentious. However, like several of my other tips (below) it concentrates on Web-based resources so it will, inevitably, date. (This is partially compensated for by there being a companion Website.)

Also, if you're coming to the subject as a computer user with a biological background, looking to exploit the many tools available, you might want to try Terry Attwood and David Parry-Smith's "Introduction to Bioinformatics" [Longman Higher Education; ISBN 0582327881], or Des Higgins and Willie Taylor's "Bioinformatics: Sequence Structure and Databanks" [Oxford University Press; ISBN 0199637903]. Another excellent practical introduction is Andreas Baxevanis and Francis Oulette's "Bioinformatics: A Practical Guide to the Analysis of Genes and Proteins" [Wiley-Interscience; ISBN 0471383910], now in its new and improved second edition. Bax teaches bioinformatics all over Canada and the experience shows.

Bioinformatics.Org also recommends Cynthia Gibas and Per Jambeck's "Developing Bioinformatics Skills" [O'Reilly, 2001 ISBN 1-56592-664-1].

Stuart Brown recommends his own book "Bioinformatics: A Biologist's Guide to Biocomputing and the Internet" [Eaton Pub Co; ISBN: 188129918X]. If he sends me a review copy I might recommend it too ;-) .

Fiction books

"Darwin's Radio" by Greg Bear [Ballantine Books, ISBN: 0345435249] is a wonderful hard SF thriller which stretches ideas derived from genome discoveries to their breaking point. It's gripping and humane.

Leonard Crane, the author of Ninth Day of Creation kindly sent me a copy for review. So far it's an excellent read. I haven't finished it yet, not because it isn't a rattling good story, but because, like "Darwin's Radio", it is very long and because I am very busy. If you'd like to read a well-researched, but speculative, novel containing actual scenes of practising bioinformatics then try it.

Ken Allen contributed the following reviews:

"Frameshift [Tor Books, ISBN: 0812571088] by Robert J. Sawyer---based around the HGP---reasonable read, but poor / confused ending."

Calculating God [Tor Books, ISBN: 0812580354]by the same author---has a subtler bio connection and is a much better read. Near the start an alien spacecraft lands, the alien emerges and says 'take me to your paleontologist'

Other lists of bioinformatics books

See also's list, Steve Brenner's list, and Aik Choon Tan's collection of books.

Centres of Bioinformatics Activity: Where is bioinformatics done?

The biggest and best source of bioinformatics links I have encountered is the Genome Web at the Rosalind Franklin Centre for Genomics Research at the Genome Campus near Cambridge, UK. Most of the links below come from that resource. My list is necessarily limited by comparison.

Research Centers

Sequencing Centers

Standard Centers

Virtual Centers for Bioinformatics Activity

Online Resources: What bioinformatics Websites are there?





Collections of Tools



Education: Where can I study Bioinformatics...

This section is not complete, but contributions to broaden its coverage are welcome. Please do not direct questions about eligibility, course quality or admissions policy to me, but to ask the individual institutions directly. Use the links to obtain contact details. If an institution doesn't provide telephone numbers/email addresses or snailmail details on its Web site it doesn't deserve your patronage.

This resource focuses on complete, full-time degree programmes rather than on individual study modules. Curating a list of the latter would be a full-time job. You can go to other places, however, if you are looking for short courses. Thanks to various [#acknowledgementsLinks contributors], including Wentian Li who pointed me to this list at Rockefeller which is mirrored at various other sites. And to Humberto Ortiz Zuazaga for mailing me a link to the ICSB, where you can find this list.

If you are interested in U.S. programmes, here's a list from Curtin and here's a list from Stanford. Thanks to Amelie Stein who also supplied some of the individual entries in this section.

Those wanting to find programmes in the Asia Pacific region could have a look at this resource maintained by the Asia Pacific Bioinformatics Network APBioNet. Thanks to Sentausa.

In the UK The Bioinformatics Resource (part of the BBSRC's CCP11 project) project maintains (among many other resources) lists of (mainly) British Masters and PhDs in bioinformatics. If you have any suggestions or updates please [/sendmessage.php? contact] me with them. You can publicize your course and offer a public service at the same time.


The Americas




Distance or Correspondence Courses

Careers: How can I become a bioinformatician?

Getting Involved

Bioinformatics Jobs

Practical tips

This section includes some simple rules-of-thumb to apply when performing common bioinformatics tasks. I try to give a reference to a more detailed source of guidance where I know of one.

Finding a Sequence

Aligning Two Sequences

Predicting the Fuctions of a Gene

How can I predict the structure of a sequence?

You could start with anyone of these excellent guides (listed strictly in alphabetical order):

How can I simulate a biomolecule?

Here's Peter J. Steinbach's "Introduction to Macromolecular Simulation"

How can I write up?

Go here to download some detailed advice. Go here for more links.

Glossary of bioinformatics terms

Here I attempt to define some common terms in bioinformatics. I have tried to balance clarity, brevity and rigour. Let me know if I let one of these priorities over-ride the others.

What is an alignment?

When two symbolic representations of DNA or protein sequences are arranged next to one another so that their most similar elements are juxtaposed they are said to be aligned. Many bioinformatics tasks depend upon successful alignments. Alignments are conventionally shown as a traces.

In a symbolic sequence each base or residue monomer in each sequence is represented by a letter. The convention is to print the single-letter codes for the constituent monomers in order in a fixed font (from the N-most to C-most end of the protein sequence in question or from 5' to 3' of a nucleic acid molecule). This is based on the assumption that the combined monomers evenly spaced along the single dimension of the molecule's primary structure. From now on I shall refer to an alignment of two protein sequences.

Every element in a trace is either a match or a gap. Where a residue in one of two aligned sequences is identical to its counterpart in the other the corresponding amino-acid letter codes in the two sequences are vertically aligned in the trace: a match. When a residue in one sequence seems to have been deleted since the assumed divergence of the sequence from its counterpart, its "absence" is labelled by a dash in the derived sequence. When a residue appears to have been inserted to produce a longer sequence a dash appears opposite in the unaugmented sequence. Since these dashes represent "gaps" in one or other sequence, the action of inserting such spacers is known as gapping.

A deletion in one sequence is symmetric with an insertion in the other. When one sequence is gapped relative to another a deletion in sequence a can be seen as an insertion in sequence b. Indeed, the two types of mutation are referred to together as indels. If we imagine that at some point one of the sequences was identical to its primitive homologue, then a trace can represent the three ways divergence could occur (at that point).

Biological interpretation of an alignment

A trace can represent a substitution:


A trace can represent a deletion:


A trace can represent a insertion:


For obvious reasons I do not represent a silent mutation.

Traces may represent recent genetic changes which obscure older changes. Here I have only represented point mutations for simplicity. Actual mutations often insert or delete several residues.

What is a DNA array?

Thanks to Bioinformatics.Org member Ravi Jain for the following answer, which I present verbatim.

DNA microarrays consist of thousands of immobilized DNA sequences present on a miniaturized surface the size of a business card or less. Arrays are used to analyze a sample for the presence of gene variations or mutations (genotyping), or for patterns of gene expression, performing the equivalent of ca. 5 000 to 10 000 individual "test tube" experiments in approximately two days of time.

Robotic technology is employed in the preparation of most arrays. The DNA sequences are bound to a surface such as a nylon membrane or glass slide at precisely defined locations on a grid. Using an alternate method, some arrays are produced using laser lithographic processes and are referred to as biochips or gene chips. The composition of DNA on the arrays is of two general types:

DNA samples are prepared from the cells or tissues of interest. For genotyping analysis, the sample is genomic DNA. For expression analysis, the sample is cDNA, DNA copies of RNA. The DNA samples are tagged with a radioactive or fluorescent label and applied to the array. Single stranded DNA will bind to a complementary strand of DNA. At positions on the array where the immobilized DNA recognizes a complementary DNA in the sample, binding or hybridization occurs. The labeled sample DNA marks the exact positions on the array where binding occurs, allowing automatic detection. The output consists of a list of hybridization events, indicating the presence or the relative abundance of specific DNA sequences that are present in the sample.

What is a homologue?

"Homology" is a much-misused term and existed in biology long before the notion of protein sequences. Strictly homology cannot be qualified; it is not correct to state that two proteins are "30% homologous" with each other, for example. If we could look back far enough in the evolutionary histories of any two molecules under comparison, we would be guaranteed to find a common ancestor eventually, but this is not true homology. An example of this would be the relationship between two variants of a single ancestral enzyme resulting from a gene duplication event.

As a rule-of-thumb, true homology should be assigned only when the feature which leads us to suspect a relationship between molecules is one we consider likely to have derived from the molecules' common ancestor. To quote Page and Holmes [Molecular Evolution: A Phylogenetic Approac, Roderick D. M. Page and Edward C. Holmes; Blackwell Scientific; ISBN 0865428891]:
"The classic molecular example is the parallel evolution of amino acid sequences in the lysozyme enzyme in leaf-eating langur monkeys and in cows. Both animals have independently evolved foregut fermentation using bacteria, and in both cases lysozyme has been recruited to degrade these bacteria. Therefore, langur and cow lysozymes are homologous as genes; however, as digestive enzymes they are not homologous because this functionality was not present in the ancestral lysozyme"
Although sequence determines structure, it is possible for two proteins to have very different sequences and functions and share a common fold. In fact, most gene products with similar three-dimensional structures are insufficiently similar at the sequence level for true homology or analogy (non-homologous similarity) to be distinguished.

What is an ontology?

Biology is changing from being a descriptive to an analytical science. Accurate and consistent descriptions are, however, vital to analysis. The idea of ontologies has been co-opted from philosophy and artificial intelligence to partition bioinformatic knowledge in a way which can be reliably navigated by computers.

[resources/Holloway02.pdf This preprint] of a review by Ele Holloway of the European Bioinformatics Institute gives a more detailed insight into the varied approaches to ontologies in bioinformatics by covering a recent meeting on the subject. The final version appears in Comparative and Functional Genomics.

What is a scoring matrix?

The following explanation was edited from a contribution by Amelie Stein.

The aim of a sequence alignment, is to match "the most similar elements" of two sequences. This similarity must be evaluated somehow. For example, consider the following two alignments:

{| summary="illustration of an alignment" |- align="center" | align="center" | (a) AIWQH AL-QH | align="center" | (b) AIWQH A-LQH |}

They seem quite similar: both contain one "indel" and one substitution, just at different positions. However, if we think of the letters as amino acid residues rather than elements of strings, alignment (a) is the better one, because isoleucine (I) and leucine (L) are similar sidechains, while tryptophan (W) has a very different structure. This is a physico-chemical measure; we might prefer these days to say that leucine simply substitutes for isoleucine more frequently---without giving an underlying "reason" for this observation.

However we explain it, it is much more likely that a mutation changed I into L and that W was lost, as in (a), than that W changed into L and I was lost. We would expect that a change from I to L would not affect the function as much as a mutation from W to L---but this deserves its own topic.

To quantify the similarity achieved by an alignment, scoring matrices are used: they contain a value for each possible substitution, and the alignment score is the sum of the matrix's entries for each aligned amino acid pair. For gaps (indels), a special gap score is necessary---a very simple one is just to add a constant penalty score for each indel. The optimal alignment is the one which maximizes the alignment score.

PAM matrices are a common family of score matrices. PAM stands for Percent Accepted Mutations, where "accepted" means that the mutation has been adopted by the sequence in question. Thus, using the PAM 250 scoring matrix means that about 250 mutations per 100 amino acids may have happened, while with PAM 10 only 10 mutations per 100 amino acids are assumed, so that only very similar sequences will reach useful alignment scores.

PAM matrices contain positive and negative values: if the alignment score is greater than zero, the sequences are considered to be related (they are similar with respect to the used scoring matrix), if the score is negative, it is assumed that they are not related. "Relationship" here may refer to evolution as well as functionality of the proteins, and of course the choice of the matrix affects the result, so one has to make an assumption on the similarity of the sequences in order to receive a useful result: rather distant sequences won't produce a good alignment using PAM 10, and the optimal aligment of two very similar sequences with PAM 500 may be less useful than that with PAM 50.

Finally, it should be noted that only some scoring matrices use similarity to evaluate alignments, but others use distance, so the be careful interpreting the results!

After this brief and necessarily superficial overview, you might want to read some more about scoring matrices.

Personal tools
wiki navigation