Bioinformatics growth curves

1
Bioinformatics growth curves

• Medline records
• Computer power
• DNA sequences
• 3-D structures

2
Database searching

• Text or sequence?
• Syntax
• Results
• Links to other resources

3
Problems

• Amount of data
• Different formats
• Lack of annotation
• Links between databases
• Quality (e.g. unfinished sequence)
• Keeping up to date

4
DNA and protein databases

• Genbank (DNA)
• EMBL (DNA)
• Swissprot - fully annotated, so delays in
  things appearing
• SCOP - Structural Classification of Proteins
• OWL - non-redundant protein database - can be
  found at HGMP website

5
Search concepts

• Boolean search - combine terms with AND, NOT, OR.
  Default is usually AND.
• Broadening or narrowing search if too few or too
  many results
• Proximity search - place quotes round words to
  specify a phrase
• Wild cards. Hemo finds hemoglobin, hemophilus,
  etc.

6
Entrez

• Available from NCBI
• Easy to use
• Sequence databases, Medline, taxonomy
• Batch mode to download and work on large datasets
• Search, review results, add terms to refine
  search

7
Sequence retrieval system (SRS)

• Available from EBI
• Interface to gt80 databases
• Searching/saving results
• Links to other databases and launch Blast,
  ClustalW
• Online help for advanced features
• Very powerful, needs learning

8
Medical genetics databases

• Human Gene Mutation
• Available at UWCM
• Mutations in human genes (mostly disease-related)

• OMIM
• Available at HGMP
• Easy to use - like a big searchable book
• Definitive clinical genetic information

9
Genome sequence browsers

• Several labs and websites, different formats
• Example in UK is Ensembl
• The Golden Path (UCSC) is my favourite
• All human (and other) genome sequence, annotated
• Easy to get various types of map
• Various ways of searching the database