Trends in Bioinformatics

1
Trends in Bioinformatics Computational
Biology(From Genomics To Proteomics)

• Professor K. Sundaram, Ph.D.
• Former Professor Head
• Department of Crystallography Biophysics
• University of Madras

2
Outline of Presentation

• What is Bioinformatics ?
• Why sudden spurt of interest recently ?
• Scope for Services/Products/Solutions ?
• What activities are possible ?
• By Individual -- By Institution -- By
  Industry
• By State -- By Country

3
What is Bioinformatics ?

• The content of this exemplary online course
  covers major areas of Bioinformatics (as of early
  1990s)
• Introduction to Biotechnology
• Pairwise Sequence Alignments
• Networking
• Multiple Alignment
• Molecular Phylogenetics
• Protein Folding

4
Birds Eye View of Biotechnology

• Every Bioinformatician must have some knowledge
  of the fields of Molecular Biology and
  Biotechnology
• An effective and succinct coverage through
  annotated illustrations has been created by
  Access Excellence
• Other such resources are also available on the
  Net
• Efforts like these (that organizations or
  individuals can undertake) are also valuable
  contributions to Bioinformatics and Biotechnology

5
Computer Technology in Biotechnology

• Computer hardware and software tools are
  integral parts of most biotechnological
  procedures - see e.g., this review of 1997
• The recent spurt in bioinformatics has resulted
  from the sequencing of all the genes in the
  entire human genome (see The Human Genome
  Project)
• The whole genome of the following species have
  also been sequenced e-coli, yeast, mouse,
  mosquito, rice
• Many others are in the pipeline
• IT is indispensable if all these data have to be
  preserved and made available to researchers

6
Analysis of cDNA Microarray Data

• Decoding of several genomes has spurred many new
  areas of intense research activity
• One such is cDNA microarray technology
• Biostatisticians and software professionals are
  sought after for the analysis of cDNA microarray
  data
• Here is a sample of theory used to normalise cDNA
  microarray data
• Theoretical prediction of the three dimensional
  structure of a protein, given its sequence, is
  another area that has attracted renewed interest
  in the postgenomic era

7
Protein Tertiary Structure Prediction

• Completion of the HGP, an enormous achievement,
  marks none the less, just the biginning of a
  bigger challenge -- The quest to decode the
  Proteome
• The base sequences in genes code for all the
  proteins.
• Proteins are responsible for virtually all the
  functions of a living system
• Activity of a protein is determined by its 3D
  structure
• There is a strong belief that the native 3D
  structure of a protein is determined by its
  primary sequence (and hence the base sequence in
  the corresponding gene)
• But the rules of the game are not yet fully
  understood

8
Ramachandran Plot

• G.N. Ramachandran is credited with the discovery
  of a major first step in Protein Structure
  Prediction
• The disposition of adjacent peptide units is
  restricted to a few allowed regions as shown in
  the adjoining Ramachandran (f, y) plot

9
PDB -- The Protein Data Bank

• The PDB is a rather unique database, much sought
  after by protein structure predictors
• Each PDB entry consists basically of the (x, y,
  z) coordinates of atoms in a protein determined
  experimentally (by x-ray crystallography or nmr)
• Experimental determination of structure requires
  the protein in very pure form (near 100)
• This was very difficult to achieve in early days
  but easier now, thanks to recombinant DNA
  technology
• Similarly methods of structure solution have
  improved significantly due to advances in
  technology

10
PDB -- The Protein Data Bank (Contd.)

• The PDB currently holds over 19,500 entries and
  the number keeps increasing rapidly
• All the same, there is no hope of determining the
  3D structures of all the proteins of the proteome
• This is the reason why 3D structure prediction
  methods have assumed a new significance
• Various techniques are being explored to work out
  structures theoretically based on patterns
  observed in PDB proteins or ab initio using
  molecular physics

11
Protein Structure Prediction

• CASP (Critical Assessment of Techniques for
  Protein Structure Prediction) is a competition
  organised every two years and makes an objective
  assessment of various prediction methods
• CAFASP is a related activity where fully
  automated protein structure prediction servers
  are assessed
• Organizers of CASP and CAFASP also conduct a
  community wide cooperative program, TMW, in an
  attempt to determine theoretically the structure
  of Ten Most Wanted proteins that are needed to
  solve scientific/ medical problems and for which
  there exists no hope of obtaining the structure
  experimentally in the near future

12
Protein Structure Prediction (Contd.)

• Present day PCs have reached a stage that
  enables an individual to participate in these
  ventures effectively
• The algorithm of such a program used to
  participate in the TMW effort is described below
• An internal coordinates representation is used
  and standard planar geometry used for the peptide
  unit. With these assumptions a model of the
  protein can be simulated given its primary
  sequence
• Cartesian coordinates of all the atoms are
  obtained looking at the molecule as a tree-like
  structure and moving from terminal atoms (leaves)
  towards the root atom and applying an affine
  transformation for each bond traversal (see next
  slide)

13
(No Transcript)
14
Protein Structure Prediction (Contd.)

• Starting conformation of a protein can be input
  in one of two forms
• (1) As a sequence of single letter aminoacid
  codes and the local dipeptide conformations at
  each a carbon atom (see example)
• (2) As a sequence of records, one for each
  residue. The record for each residue contains the
  residue name followed by the values of variable
  torsion angles f, y, and, cs (see example)
• Total nonbonded energy is calculated using these
  formulae

15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
Protein Structure Prediction (Contd.)

• Various strategies are used to generate likely
  conformations for the protein using a simulated
  annealing approach adopting the Monte Carlo
  algorithm
• A more detailed description is available in this
  article

19
(No Transcript)
20
Proteomics Research

• Once the structure of a protein is obtained,
  further research can be carried out in areas like
  functional genomics and proteomics
• Several proteins are suspected to engage in
  function related specific associations. This can
  be investigated theoretically as discussed in
  this article