Bioinformatics

M.Alroy Mascrenghe 1
Introduction……

2000
n  A Major event happened that was to
change the course of human history
n  It was a joint British and American
effort
n  nothing to do with IRAQ!
n  It was a race – who will complete first
n  Race Test – not whether they have
taken drugs but whether they can
produce them!
n  Human genome was sequenced

A Situ…somewhere in the
near future
n  A virus –not ‘I love you’ virus- creates an epidemic
n  Geneticists and bioinformaticians role on their
sleeves
n  Genetic material of the virus is compared with the
existing base of known genetic material of other
viruses
n  As the characteristics of the other viruses are
known
n  From genetic material computer programs will
derive the proteins necessary for the survival of the
virus
n  When the protein (sequence and structure) is
known then medicines can be designed

What is
n  The marriage between computer
science and molecular biology
l  The algorithm and techniques of
computer science are being used to
solve the problems faced by molecular
biologists
n  ‘Information technology applied to
the management and analysis of
biological data’
l  Storage and Analysis are two of the
important functions – bioinformaticians
build tools for each

Biology Chemistry
Statistics
Computer
Science
Bioinformatics

What is..
n  This is the age of the Information
Technology
n  However storing info is nothing new
n  Information to the volume of
Britannica Encyclopedia is stored in
each of our cells
n  ‘Bioinformatics tries to determine
what info is biologically important’

Basics
of
Molecular Biology….

DNA & Genes
n  DNA is where the genetic information is
stored
n  Blonde hair and blue eyes are inherited by
this
n  Gene - The basic unit of heredity
l  There are genes for characteristics i.e. a gene
for blond hair etc
n  Genes contain the information as a
sequence of nucleotides
n  Genes are abstract concepts – like
longitude and latitudes in the sense that
you cannot see them separately
n  Genes are made up of nucleotides

Nucleotide (nt)
n  Each nt I made up of
l  Sugar
l  Phospate group
l  Base
n  The base it (nt) contains makes the only
difference between one nt and the other
n  There are 4 different bases
n  G(uanine),A(denine),T(hymine),C(ytosine)
n  The information is in the order of nucleotide
and the order is the info
n  Genes can be many thousands of nt long
n  The complete set of genetic instructions is
called genomes

Chromosomes
n  DNA strings make
chromosomes
n  Analogy
l  Letters - nt
l  Sentences – genes
l  Individual volumes of Britannica
encyclopedia – chromosomes
l  All voles together - Genome

Double Helix
n  The DNA is a double helix
n  Each strand has complementary
information
n  Each particular base in one strand is
bonded with another particular base in the
next strand
l  G - C
l  A - T
n  For example -
l  AATGC one strand
l  TTACG other strand

Proteins
n  Proteins are very important
biological feature
n  Amino Acids make up the proteins
n  20 different amino acids are there
n  The function of a protein is
dependant on the order of the amino
acids

Proteins…
n  The information required to make aa is
stored in DNA
n  DNA sequence determines amino acid
sequence
n  Amino Acid sequence determines protein
structure
n  Protein structure determines protein
function
n  A Substance called RNA is used to carry
the Info stored in the DNA that in turn is
used to make proteins
n  Storage - DNA
n  Information Transfer – RNA
n  RNA is the message boy!

Central dogma
DNA transcription RNA Translation Protein
RNA Polymerase Ribosomes

Proteins…..
n  Since there are 20 amino acids to
translate one nt cannot correspond
to one aa, neither can it correspond
as twos
n  So in triplet codes – codon – protein
information is carried
n  The codons that do not correspond
to a protein are stop codons – UAA,
UAG, UGA (RNA has U instead of T)
n  Some codons are used as start
codons - AUG as well as to code
methionine

Protein Structure
n  Shows a wide variety as opposed to the
DNA whose structure is uniform
n  X-ray crystallography or Nuclear Magnetic
Resonance (NMR) is used to figure out the
structure
n  Structure is related to the function or rather
structure determines the function
n  Although proteins are created as a linear
structure of aa chain they fold into 3 d
structure.
n  If you stretch them and leave them they will
go back to this structure – this is the native
structure of a protein
n  Only in the native structure the proteins
functions well
n  Even after the translation is over protein
goes through some changes to its structure

Gene Expression
n  Gene Expression – the process of
Transcripting a DNA and translating a RNA
to make protein
n  Where do the genes begin in a
chromosome?
n  How does the RNA identify the beginning
of a gene to make a protein
n  A single nt cannot be taken to point out the
beginning of a gene as they occur
frequently
n  But a particular combination of a nucleotide
can be
n  Promoter sequences – the order of nt
which mark the beginning of a gene

Bioinformatics
Techniques…..

Prediction and Pattern
Recognition
n  The two main areas of bioinformatics
are
n  Pattern recognition
l  ‘A particular sequence or structure has
been seen before’ and that a particular
characteristic can be associated with it
n  Prediction
l  From a sequence (what we know) we
can predict the structure and function
(what we don’t know)

Dot plots….
n  Simple way of evaluating
similarity between two
sequences
n  In a graph one sequence is on
one side the next on the other
side
n  Where there are matches
between the two sequences the
graph is marked

Alignments
n  A match for similarity between the characters of two or
more sequences
n  Eg.
l  TTACTATA
l  TAGATA
n  There are so many ways to align the above two
sequences
l  1.
n  TTACTATA
n  TAGATA
l  2.
n  TTACTATA
n  TAGATA
l  3.
n  TTACTATA
n  TAGATA
n  So which one do we choose and on what basis?
n  Solution is to Provide a match score and mismatch score

Gaps
n  Introduce gaps and a penalty
score for gaps
n  TTACTATA
n  T_A_GATA
n  In gap scores a single indel which is two characters long is preferred to two indels which are each one
character long
n  However not all gaps are bad
l  TTGCAATCT
l  CAA
l  How do we align?
l  ---CAA---
l  These gaps are not biologically significant
l  Semi Global Alignments

Scoring Matrix
n  For DNA/protein sequence alignment we create a matrix
n  If A and A score is 1
n  If A and T score is -5
n  If A and C score is -1

Dynamic Programming
n  As the length of the query sequences
increase and the difference of length
between the two sequence also increases
–more gaps has to be inserted in various
places
n  We cannot perform an exhaustive search
n  Combinatorial explosion occurs – too much
combinations to search for
n  Dynamic programming is a way of using
heuristics to search in the most promising
path

Databases
n  Sequence info is stored in
databases
n  So that they can be manipulated
easily
n  The db (next slide) are located
at diff places
n  They exchange info on a daily
basis so that they are up-to-date
and are in sync
n  Primary db – sequence data

Major Primary DB
Nucleic Acid Protein
EMBL (Europe) PIR -
Protein Information
Resource
GenBank (USA) MIPS
DDBJ (Japan) SWISS-PROT
University of Geneva,
now with EBI
TrEMBL
A supplement to SWISS-
PROT
NRL-3D

Composite DB
n  As there are many db which one to
search? Some are good in some
aspects and weak in others?
n  Composite db is the answer – which
has several db for its base data
n  Search on these db is indexed and
streamlined so that the same stored
sequence is not searched twice in
different db

Composite DB
n  OWL has these as their primary
db
l  SWISS PROT (top priority)
l  PIR
l  GenBank
l  NRL-3D

Secondary db
n  Store secondary structure info or
results of searches of the
primary db
Compo
DB
Primary
Source
PROSITE SWISS-PROT
PRINTS OWL

Database Searches
n  We have sequenced and identified
genes. So we know what they do
n  The sequences are stored in
databases
n  So if we find a new gene in the
human genome we compare it with
the already found genes which are
stored in the databases.
n  Since there are large number of
databases we cannot do sequence
alignment for each and every
sequence
n  So heuristics must be used again.

Areas in
Bioinformatics…

Genomics
n  Because of the multicellular structure, each
cell type does gene expression in a
different way –although each cell has the
same content as far as the genetic
n  i.e. All the information for a liver cell to be a
liver cell is also present on nose cell, so
gene expression is the only thing that
differentiates

Genomics - Finding Genes
n  Gene in sequence data – needle in a
haystack
n  However as the needle is different
from the haystack genes are not diff
from the rest of the sequence data
n  Is whole array of nt we try to find and
border mark a set o nt as a gene
n  This is one of the challenges of
bioinformatics
n  Neural networks and dynamic
programming are being employed

Organism Genome
Size
(Mb)
bp * 1,000,000
Gene
Number
Web Site
Yeast 13.5 6,241 http://genome-
www.stanford.ed
u/
Saccharomyces
Fruit Flies 180 13,601 http://
flybase.bio.india
na.edu
Homo
Sapiens
3,000 45,000 http://
www.ncbi.nlm.ni
h.gov/genome/

Proteomics
n  Proteome is the sum total of an
organisms proteins
n  More difficult than genomics
l  4 20
l  Simple chemical makeup complex
l  Can duplicate can’t
n  We are entering into the ‘post
genome era’
n  Meaning much has been done with
the Genes – not that it’s a over

Proteomics…..
n  The relationship between the RNA and the protein it codes are
usually very different
n  After translation proteins do change
l  So aa sequence do not tell anything about the post
translation changes
n  Proteins are not active until they are combined into a larger
complex or moved to a relevant location inside or outside the cell
n  So aa only hint in these things
n  Also proteins must be handled more carefully in labs as they tend
to change when in touch with an inappropriate material

Protein Structure Prediction
n  Is one of the biggest challenges
of bioinformatics and esp.
biochemistry
n  No algorithm is there now to
consistently predict the structure
of proteins

Structure Prediction methods
n  Comparative Modeling
l  Target proteins structure is
compared with related proteins
l  Proteins with similar sequences
are searched for structures

Phylogenetics
n  The taxonomical system reflects
evolutionary relationships
n  Phylogenetics trees are things which reflect
the evolutionary relationship thru a picture/
graph
n  Rooted trees where there is only one
ancestor
n  Un rooted trees just showing the
relationship
n  Phylogenetic tree reconstruction algorithms
are also an area of research

Applications….

Medical Implications
n  Pharmacogenomics
l  Not all drugs work on all patients, some good
drugs cause death in some patients
l  So by doing a gene analysis before the
treatment the offensive drugs can be avoided
l  Also drugs which cause death to most can be
used on a minority to whose genes that drug is
well suited – volunteers wanted!
l  Customized treatment
n  Gene Therapy
l  Replace or supply the defective or missing gene
l  E.g: Insulin and Factor VIII or Haemophilia
n  BioWeapons (??)

Diagnosis of Disease
n  Diagnosis of disease
l  Identification of genes which cause the
disease will help detect disease at early
stage e.g. Huntington disease -
n  Symptoms – uncontrollable dance like
movements, mental disturbance,
personality changes and intellectual
impairment
n  Death in 10-15 years
n  The gene responsible for the disease has
been identified
n  Contains excessively repeated sections of
CAG
n  So once analyzed the couple can be
counseled

Drug Design
n  Can go up to 15yrs and
$700million
n  One of the goals of
bioinformatics is to reduce the
time and cost involved with it.
n  The process
l  Discovery
n  Computational methods can
improves this
l  Testing

Discovery
Target identification
l  Identifying the molecule on which the
germs relies for its survival
l  Then we develop another molecule
i.e. drug which will bind to the target
l  So the germ will not be able to interact
with the target.
l  Proteins are the most common targets

Discovery…
n  For example HIV produces HIV
protease which is a protein and
which in turn eat other proteins
n  This HIV protease has an active
site where it binds to other
molecules
n  So HIV drug will go and bind
with that active site
l  Easily said than done!

Discovery…
n  Lead compounds are the
molecules that go and bind to
the target protein’s active site
n  Traditionally this has been a trial
and error method
n  Now this is being moved into the
realm of computers

Related Computer
Technology………….

PERL
n  Perl is commonly used for
bioinformatics calculations as its
ability to manipulate character
symbols
n  The default CGI language
n  It started out as a scripting language
but has become a fully fledged
language
n  IT has everything now, even web
service support
n  http://bio.perl.org

The place of XML & Web
Services
n  Various markup languages are being created –
Gene Markup language etc to represent sequence/
gene data
n  Web Services – program to program interaction,
making the web application centric as opposed to
human centric
n  So this has to platform language independent
n  Protocols like SOAP help in this regard
n  In bioinformatics various databases are being used,
different platforms, languages etc
n  So web services helps achieve platform
independence and program interaction
n  Since sequence data bases are in various formats,
platforms SOAP also helps in this regards

The place of GRID
n  GRID - new kid on the block
n  Using many computers to fulfill a
single computational tasks
n  Bioinformatics is the ideal
platform as it has to deal with a
large amount of data in
alignment and searches
n  E-science initiative in the UK
n  ORACLE 10g – the worlds first
GRID database

Data bases and Mining
n  Lot of the sequence databases are
available publicly
n  As there is a DB involved various
data mining techniques are used to
pull the data out
n  As there is a lot of literature – articles
etc – on this area a data mining on
the literature – not on the sequence
data has also become a PhD topic
for many

European Molecular Biology
Network (EMBnet)
n  A central system for sharing, training
and centralizing up to date bio info
n  Some of the EMBnet sites are:
n  SQENET
l  http://www.seqnet.dl.ac.uk
n  UCL
l  http://www.biochem.ucl.ac.uk/bsm/
dbbrowser/embnet/
n  EBI – European Bioinformatics
Institute
l  www.ebi.ac.uk

References
n  Dan E. Krane and Michael L. Raymer
l  Basic Concepts of Bioinformatics
n  Arthur M Lesk
l  Intro to Bioinformatics
n  T.K. Attwood & D. J. Parry-Smith
l  Intro to Bioinformatics
n  The genetic Revolution
l  Dr Patrick Dixon
n  Prof David Gilbert’s Site
l  http://www.brc.dcs.gla.ac.uk/~drg/

Thank You!

Bioinformatics

More Related Content

What's hot (20)

Similar to Bioinformatics (20)

More from Bahauddin Zakariya University lahore (20)

Recently uploaded (20)

Bioinformatics