Theory

• The first thing you will need in an analysis are some data!
• These may come from a variety of sources, and in a variety of formats

Public databases

• General databases, part of the International Nucleotide Sequence Database Collaboration
  • GenBank
  • European Nucleotide Archive (ENA)
  • DNA Databank of Japan (DDBJ)
• Virus specific databases, e.g.
  • NCBI Influenza Virus Resource
  • Los Alamos HIV Sequence Database

Genbank

• We will download sequences and associated data from GenBank
• Central to this goal is the way in which sequences are identified
• GI numbers are simply a series of digits that are assigned consecutively
• Accession numbers in the nucleotide database comprise of 1-2 letters, 5-6 digits, a dot and a version number
• In the reference sequence database, accession numbers comprise of 2 letters, an underscore, and several digits

Annotations

• At a minimum, GenBank expects the following for each viral sequence submitted
  • Organism name
  • Isolate
  • Country (optional)
  • Collection date (optional)
  • Host (optional)
• For influenza virus, rotavirus, and norovirus, the above optional fields are mandatory

Entrez through the web

• Multiple databases at NCBI can be searched via the search engine Entrez
• For exploratory searches, it is useful to access Entrez directly through a web browser at http://www.ncbi.nlm.nih.gov/Entrez

Output formats

• There are a number of different formats for output once sequences have been identified; we will consider three:
  • FASTA
  • List of GI numbers
  • INSDSeq XML

Entrez through R

• Eutils can also be used to access Entrez
• This has the advantages of being able to reproduce the searches, as well as update them, as new data become available

Extracting annotation data

• Information like host and collection data is not provided in the FASTA file; it has to be extracted from the full record
• In the practical, we will see how this information can be extracted from the record in INSDSeq XML format