• View
• Edit
• History
• Print

IntroductionToBioinformaticsOlder

Introduction to Bioinformatics

Aidan Budd and Venkata Satagopam

We begin by exploring some common features of bioinformatics tools and analyses - this will help provide a framework for understanding and criticising your analyses.

1. What does "Bioinformatics" mean?

• Bioinformatics: the analysis of biological data using computers

2. Bioinformatics involves many different types of data

• Protein Sequence
• Nucleotide Sequence
• Phylogenetic Trees
• DNA Microarray Data
• Multiple Sequence Alignments
• Biological Literature
• 3D Molecular Structures e.g. proteins
• Mass Spectrometry Data
• etc.

3. ALL bioinformatic applications depend on identifying similarity between datasets

4. ALL bioinformatic tools are designed to address one or both of these questions:

Exercises - Recognising features of bioinformatic tools

In discussion with your neighbours, consider a situation (or several situations) where you have used bioinformatic tools in the past.

While using these tools:

• Which of these questions were you trying to answer:
  • "What is already known about my sequence/system of interest?"
  • "What can I predict about my sequence/system that I can use as a hypothesis to test in a wet-lab experiment?"
  • Or perhaps in your opinion you were addressing neither of these questions? If so, what is the central question that was being addressed?
• Did the tool work via identifying similarity between datasets in a similar way to that described above?
  • If not, then can you describe in similar terms how the tool functioned?
  • If yes, which were the datasets being compared? And how was similarity between the datasets being assessed?

If you have only limited experience working with bioinformatic tools, then select one (or more) of the tools described in the publications listed below and use them in this exercise.

ProtTest: selection of best-fit models of protein evolution.

Abascal et al. 2005 Bioinformatics 21(9) 2104-5 PMID: 15647292

JIGSAW: integration of multiple sources of evidence for gene prediction.

Allen and Salzberg 2005, Bioinformatics 21(18) 3596-603 PMID:16076884

IUPred: web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content.

Dosztanyi et al. 2005 Bioinformatics 21(16) 3433-4 PMID:15955779

4DXpress: a database for cross-species expression pattern comparisons.

Haudry et al. 2007 Nucleic Acids Research 36(Database issue):D847-53 PMID:17916571

5. Bioinformatics databases

Being aware of some of the typical features of bioinformatic database records/files can help us find what we need in this jungle of data

A. Primary accession numbers and cross-references

Primary accession numbers aim to provide a string/number that unambiguously identifies just one record within a given database, distinguishing it from all other records in that database.

Demonstration

We'll show you the difference between trying to find the UniProt human Src kinase record without and then with knowledge of the accession number

Exercise

To illustrate the importance of accession numbers in navigating through biological databases, attempt to find records in the following databases associated either with human hemoglobin subunit beta, or with your own protein of interest, without using accession numbers (if using the example of human hemoglobin beta, we've provided links to the relevant accession numbers for you to compare with the results of your own searches.)

Note that in some cases there may be several accession numbers referring to the same entity e.g. 1A00 and 1BBB are both PDB accession numbers for 3D structures of hemoglobin.

• UniProt Accession Number
• Ensembl Gene Accession Number
• STRING Accession Number
• EMBL Nucleotide Database Accession Number

B. Overlapping bioinformatic resources

Demonstration

Another common situation you'll find when searching bioinformatic resources is that, in many cases, several different resources provide the same/similar kinds of information. Often you will find that there is some, but not complete, overlap between the information provided by these different sites.

For example PhosphoSite, Human Protein Reference Database, Phospho.ELM, and UniProt all provide information about experimentally-demonstrated phosphorylation sites e.g. for the human epsin-1 sequence (UniProt name EPN1_HUMAN).

For this reason, if you have the time and are very interested in finding evidence of particular features of your protein, it makes sense to examine more than just one resource for information of this kind.

Exercise

Using the resources above, determine the extent of overlap in the set of sites described to be phosphorylated in the protein CBP_HUMAN (or in your own protein of interest).

C. User-friendly data visualisation

Providing representations of biological data in an easy-to-digest form is a key aspect of bioinformatics development and research.

Exercise

To give an example of the difference between the raw data, and more user-friendly representations, contrast the information found in the following links, all of which present the UniProt record for CBP_HUMAN:

• UniProt.org
• ExPASy
• Raw format (also obtained via UniProt.org)
• SRS-EMBL

For each of these sites, try and find the parts of the record that refer to:

• Domains/Families predicted in the sequence by PFAM
• GO (Gene Ontology) Accession numbers/IDs with which the record in annotated e.g. GO:0000123
• The amino-acid sequence

Is one of the versions of the record easier to use than the others?

Are there particular features of the representations that make them easier/more difficult for you to navigate through?

6. SRS

Exercise 1

• When we initially started to know about a gene, a protein or a disease, we start with a quick overview of the information related to it.

Exercise 2

• Now we want to run some slightly bigger queries with more than one search field

Exercise 3

• Now run linking query

Learning Objectives

• Outlining the scope of the field of bioinformatics
• Making participants aware of common features of many/most bioinformatics tools/analyses, and providing practice at deconstructing the tools - an important step towards critical evaluating them
• Raising awareness of some common features of biological databases