Bioinformatics + Homology Modelling

Things you need to know! • Databanks vs. databases • The difference between a primary and a secondary databank • Examples of primary and secondary databanks • The definition of homology / orthology / paralogy and their importance • The difference between a databank (like SwissProt) and a search tool (like BLAST) • Local and global alignment • Basic ideas of sequence searching and the meaning of the e-value • Problems of annotation

Define bioinformatics

Aiding the Biologist in creating, storing, searching and analyzing biological data (particularly sequences and structures, but also 'omics data), presenting it in a way Biologists can use and applying the analysis to make predictions

What are we trying to achieve with bioinformatics?

What is a key point both homologous proteins/genes share?

• Have descended from a common ancestor

• Homologues have derived from a common ancestor hence they will have similar characteristics

“It is WRONG to refer to ‘percentage homology’”, why?

• An old term used before the evolutionary angle took over the word homology

• Percentage sequence identity is used instead to infer homology

What is the criteria for a protein sequence to be always homologues? (and vice versa)

100AA with >35% sequence identity = Homologues and same function still

Lower sequence identity = May be homologues

Which source can you use to compare structures of homologues and determine which type of homologue?

Structure comparison

• PDB Database

• GPCRdb

Identifying type

• NCBI Homologene

• EBI

What are the two types of homologues?

• Orthologues

• Paralogues

What are orthologues?

Homologues resulting from a speciation event

i.e. the same gene in two different species

What are paralogues? Give an example

Result from gene copying within a species

E.g. globin genes (beta and alpha globin)

What is protein modelling, why use it, and why not just rely on software predictions?

Protein modelling = Using instrumentation to work out the actual protein structure, i.e. cryo-EM

You need to validate things experimentally to verify that what you’ve computed is correct (support the computational prediction), and you get the results you expected

What are the two types of storages of data?

Databank & database

Define databank

A collection of data (normally in simple text files) without a fixed associated query tool

• You can download all the raw data and handle it however you wish

What are the types of databanks?

Primary

• Raw sequence/structure data

• Possibly with detailed annotations

Secondary = Curates the primary data

• Derived data - sequence profiles, etc

• Generally highly annotated

Give examples of primary databanks, and what they (+may) contain

DNA: Genbank, EMBL-ENA, DDBJ

Protein: UniProtKB/SwissProt

• Simply contain sequence data (DNA or protein)

• May also have ‘feature’ information (splice sites, signal sequences, disulphides, active sites, etc., etc.)

• DNA databanks may also contain translations (known or predicted)

What is the main primary databanks for structural data and enzyme classifications (EC numbers)?

• PDB

• Enzymes

Give examples of primary databanks, and what they (+may) contain

PROSITE, PRINTS, BLOCKS, INTERPRO

• Contain derived information

• Patterns that characterize a protein family

• Detailed annotation

NOTE: Can be used for finding homologues

Where would you find annotations of proteins/genes from the source?

• References

• Methods

• Cross-links to other databases

• ‘Feature tables’

  • Segments of biological significance (e.g. coding regions, active sites)

• More general descriptions

  • Computer parsable?

  • May use a ‘restricted dictionary’ or ‘ontology’

• Authors

NOTE: Just because an information isn’t there doesn’t necessarily mean it is not known

How are most sequences predicted and why are gene-prediction methods imperfect?

• Predicted from genomic data

• Most of the the time the software only makes a sensible prediction after more biological information is put into the database

“A protein identified from genome data is hypothetical until verified by experiment”

Define database

A structured collection of data hidden behind a fixed search tool

What is protein/peptide sequencing largely replaced by?

DNA sequencing (now automated)

What are the two important things to remember about quality?

• The quality of raw data is as good as the methods that produce it

• The quality of annotations is as good as the curators

How were annotations in the pre-genome world different to annotation in the genome world?

PRE

• DNA sequence came from a single group with particular interest in that gene

• Annotations grounded in experiment

• Written by experts in that gene

POST

• Rarely experimental confirmation of expression of putative genes or characterization of products

• Annotation based on computer analysis

• The weakest link of genomics!

• Getting it right is labour intensive and underfunded

• Poor annotation negates the benefit of having genome sequence data!