Biological Databases

Dr. Upeksha Ganegoda

Department of Computational Mathematics

1
 Outline
 Different types of biological networks
 Database Structures
 Biological database types based on content

2
 Biological Networks
 Protein-protein interaction network
 Metabolic network
 Gene regulatory network
 RNA network

3
 Protein-protein interaction network
 A protein can interact with another protein, in order to build
a protein complex or to activate it. By using a protein-
protein interaction network it shows how and which proteins
are interact each other.
 Node represent a protein, Arc represent the interaction
between two protein.
 Can use different types of graph algorithms to identify:
 Protein complexes
 Protein functions
 Protein Hubs
 etc

4
 Protein Hub?

Protein complexes?

5
 Metabolic network
 Metabolic networks give an in-depth insight of the molecular
mechanisms of a particular organism. It will correlate the
genome with molecular physiology and provide the most
comprehensive of all biological networks.

Ex: Databases such as the Kyoto Encyclopedia of Genes and

Genomes (KEGG) and the Biochemical Genetic and
Genomics knowledgebase (BIGG) contain the metabolic
network of a wide range of species.

6
 Gene-regulatory network
 It is a common type of regulatory network
 gene regulatory network consist of DNA segments in a cell
which interact with each other indirectly (by using their
RNA and protein expression products) and with other
materials in the cell to manage the gene expression levels of
mRNA and proteins.

7
 RNA networks
 RNA networks show the interaction between RNA-RNA or
RNA-DNA interactions. By understanding the microRNA’s
role in disease, the researchers able to construct microRNA-
gene networks by using predicted microRNA targets
available in public databases such as Target Scan, PicTar,
microRNA, miRBase and miRDB.

8
 Representation as a network
Network G = (V, E, w), where V represents the set of proteins,
E is the set of interactions and w denotes the weight of each
interaction
Network can construct as
 Directional graph Ex: gene-regulatory graph
r1

g1

g2
 Bidirectional graph Ex: PPI network r2

p1

p1

p1
9
 Main functions of biological databases
 Make biological data available to scientists.
As much as possible of a particular type of information should
be available in one single place (book, site, database). Published
data may be difficult to find or access, and collecting it from
the literature is very time-consuming. And not all data is
actually published explicitly in an article (genome sequences).
 To make biological data available in computer-
readable form.
Since analysis of biological data almost always involves
computers, having the data in computer-readable form (rather
than printed on paper) is a necessary first step.

10
 What is a database?
 How can data be stored...

Flat-file format, with fields separated by some delimiter

Nancy|Dengler|Botany|University of Toronto|25 Willocks St, Toronto, ON. M5S 3B2
Peter|Lewis|Dept. of Biochemistry|Uni. Toronto|1 King’s College Circle, Toronto, ON. M5S 1A8
John|Coleman|Department of Botany|University of Toronto|25 Willcocks St, Toronto, ON. M5S 3B2
John|Coleman|Dept. of Biology|York University|4700 Keele St, Toronto, ON. M3J 1P3

These data could also be stored in a spreadsheet

What are the problems with this sort of database?

Relational Databases offer a solution...

11
 Database structures
 Flat files
 Relational
 Object oriented

12
 Relational database
Nancy|Dengler|Botany|University of Toronto|25 Willocks St, Toronto, ON. M5S 3B2
Peter|Lewis|Dept. of Biochemistry|Uni. Toronto|1 King’s College Circle, Toronto, ON. M5S 1A8
John|Coleman|Department of Botany|University of Toronto|25 Willcocks St, Toronto, ON. M5S 3B2
John|Coleman|Dept. of Biology|York University|4700 Keele St, Toronto, ON. M3J 1P3

A relational database consists of a relations (tables) containing attributes (fields or

columns). Each row in a table is known as a tuple or a record. Information should be
‘normalized’ so that it is non-redundant this means that every row should be unique,
although this ideal is not always observed.

Professor_id First_name Last_name Contact_id

Table 1 Nancy Dengler 1
2 Peter Lewis 2
'Professors' 3 John Coleman 1
4 John Coleman 3

Contact_id Institution Department Address

Table 1 University of Toronto Dept. of Botany 25 Willocks St, Toronto, ON. M5S 3B2
2 Uni. Toronto Dept. of Biochemisty 1 King’s College Circle, Toronto, ON. M5S 1A8
'Contacts' 3 York University Dept. of Biology 4700 Keele St, Toronto, ON. M3J 1P

13
14
 Different Database Types
 Primary databases
Contain original biological data. Ex. Raw nucleic acid sequence data from
GeneBank, EMBL database, DNA Data Bank.
 Secondary databases
Contain computationally processed or manually curated information based
on original information from primary database. Ex. SWISS-PROT, TrEMBL
(contain translated nucleic acid sequences), PIR (contain annotated protein
sequences).
 Specialized databases
 This will cater to a particular research interest. Ex. Flybase, WormBase,
AceDB, and TAIR

15
 Pitfalls of biological databases
 Overreliance of sequence information without understanding
the reliability of the information.
 High level of redundancy
 Annotations of genes can occasionally be false or incomplete.

16
 Accession codes, identifiers
 Many of the biological databases (GenBank, UNIPROT etc.)
have two (or more!) different ways of identifying a given
entry:
• Identifier
• Accession code (or number)

17
  Identifier
An identifier ("locus" in GenBank, "entry name" in UNIPROT) is a
string of letters and digits that understandable in some meaningful way
by a human.

Identifiers are not as stable as accession numbers, mainly because they are
modified by the curators if the presumed function of the protein is found
to be something else.

UNIPROT: B5YME7
GenBank: XM_002295694

An identifier can change. For example, the database curators may decide
that the identifier for an entry no longer is appropriate. This can happen
very rarely.

18
  Accession code (number)

An accession code (or number) is a number (with a few

characters in front) that uniquely identifies an entry. It is often
assigned arbitrarily. For example, the accession code for
B5YME7_THAPS in UNIPROT is B5YME7.

In the case of GenBank, the accession code for the human

BRAC2 gene sequence is XM_002295694.

19
 Versions and Gene Indices
In 1992, NCBI began assigning a unique number for each sequence
submitted – the GenInfo Identifier (GI) number. The same accession number
may be associated with a different GI if a newer or corrected sequence is
submitted.

Records typically contain the Accession.Version identifier, such as

XM_002295694.1, in the VERSION line of the record. This identifier is
mapped to its unique corresponding GI number, which is the “primary key”
of GenBank.
To specify a sequence exactly in GenBank, use either its GI or
Accession.Version. To retrieve the most up-to-date sequence, use the
accession number without version.

20
21
22
 GenBank Flatfile Format (GBFF)

 The GenBank flatfile format (GBFF) explain the nucleotide sequences of a specific
gene. It contains all of the information associated with the sequence, as well as the
sequence itself.
The GBFF has 3 parts: the header, the features, and the sequence itself.

identifier length source type NCBI entry date

taxonomic group
23
 GenBank flatfile format - Header

DEFINITION: The biology of the molecule in a sentence.

ACCESSION: Code(s)
VERSION: Number; GI number
KEYWORDS: Keywords as defined by the submitters

24
 SOURCE: Contains organism
name
ORGANISM: Contains complete
taxonomic information from the
NCBI taxonomy server.
REFERENCE: Details on a
publication about the sequence.
COMMENT: Contains misc.
information and revision details.

25
 GenBank Flatfile Format – Features

A direct representation of the biological information in the

record.
The Source Feature must be present in all GenBank records, and
contains information as to where the molecule comes from
/organism = “Homo sapiens”, and, potentially, map, chromosome
and tissue type information.
 In some records the CDS (coding sequence) feature is present:

26
27
 GenBank Flatfile Format – Sequence
 The last part of the GenBank flat file record is the sequence
itself:

28
 Nucleotide Databases – Growth of
GenBank
 from http://www.ncbi.nlm.nih.gov/genbank/statistics

29
 Other facilities in NCBI database

30
 Disease details

31
 Gene Details

32
 Gene expression details….

33
34
35
36
37
38
39