Future Challenges in Bioinformatics

1
Future Challenges in Bioinformatics
2
Introduction

• Introduction How RRX got involved
• Life sciences context How bioinformatics came to
  be important
• The past half century How bioinformatics has
  evolved

3
Introduction

• Categories of Bioinformatics Tools
• Why We Need Supercomputers
• Software Development Issues
• Future Challenges
• Tools for Biotech Projects
• Summary

4
How RRX got involved

• Submitted a Canadian Foundation for Innovation
  (CFI) proposal for Advanced Bioinformatics
  Collaborative Computing (ABioCC)

5
How RRX got involved

• Developed an SVG based visualization front end
• Paper will be presented at SVG Open 2003 in
  Vancouver on July 17th

6
How bioinformatics came to be important

• After the structure of DNA was reverse engineered
  with X-Ray diffraction in 1953 focus shifted to
  nucleic acid sequence analysis
• DNA/RNA/protein sequence data accumulated using
  computer programs for storage and analysis

7
How bioinformatics came to be important

• Bioinformatics algorithms in development for the
  last half century came into wide spread use by
  researchers
• The ability to compare sequences created a
  homology context for unknown sequences of
  interest leading to advances

8
How bioinformatics came to be important

• Improved sequencing technology enabled the
  complete deciphering of the human genome gtgtgt 1999
• About 3.18 billion base pairs
• Celera used 300 PE Biosystems ABI Prism 3700 DNA
  Analysers

9
How bioinformatics has evolved

• Central dogma of molecular biology
• DNA sequences are transcribed into mRNA
  sequences, mRNA sequences are translated into
  protein sequences, which fold 3D creating
  structures with functions statistically survival
  selected gtgtgt affecting the prevalence of the
  underlying DNA sequences in a population

10
How bioinformatics has evolved

• This created a supporting information flow
• Organization and control of genes in the DNA
  sequence
• Identification of transcriptional units in the
  DNA sequence
• Prediction of protein structure from sequence
• Analysis of molecular function

11
How bioinformatics has evolved

• Another covariant information flow was created
  based on the scientific method
• Create hypothesis wrt biological activity
• Design experiments to test the hypothesis
• Evaluate resulting data for compatibility with
  the hypothesis
• Extend/modify hypothesis in response

12
How bioinformatics has evolved

• IT used to handle explosion of data from high
  throughput techniques, too complex for manual
  analysis
• X-ray diffraction

13
How bioinformatics has evolved

• Automated DNA sequencing
• Amersham Biosciences
• Applied Biosystems
• Beckman Coulter
• LI-COR
• SpectruMedix Corp.
• Visible Genetics Corp.

14
How bioinformatics has evolved

• Microarray expression analysis

15
How bioinformatics has evolved

• Rapid emergence of 3D macromolecular structure
  databases
• New sub discipline structural bioinformatics
• Atomic and sub cellular spatial scales
• Representation/physics
• Storage/retrieval/source data correlation/interpre
  tation
• Analysis/simulation
• Display/visualization

16
How bioinformatics has evolved
17
Categories of Bioinformatics Tools

• Databases gtgtgt search/compare
• Sequence Analysis - Clusters
• Genomics
• Phylogenics
• Structure Prediction
• Molecular Modelling
• Microarrays
• Packages, Misc Apps, Graphics, Scripts

18
Categories of Bioinformatics Tools

• Database gtgtgt search/compare

• aceperl
• BLAST
• Blastall
• Blastpgp
• BLAT
• Blimps
• Entrez
• FASTA
• fastacmd
• formatdb
• getz

• HMMER
• IMPALA
• InterProScan
• PHI-BLAST
• ProSearch
• PSI-BLAST
• PSI-BLASTN
• Seguin
• Swat
• tace
• xace

19
Sequence Analysis

• Artemis
• Bl2seq
• BLAST
• Clustal W, X
• consed/autofinish
• Cross_match
• Dotter
• EMBOSS
• FASTA
• Glimmer
• HMMER
• InterProScan
• MEME
• View

• Paracel Transcript Assem
• Phrap
• Phred
• Primers
• ProSearch
• Readseq2
• Rnabob
• RRTree
• SAPS
• seals
• Seqsblast
• STADEN
• Swat
• T-Coffee

20
Genomics

• Calc_primers
• Cross_match
• FPC
• GENSCAN
• Glimmer
• Image
• Mzef
• Phrap

• Phred
• STADEN
• Swat
• tace
• tace_celegans
• tRNAscan-SE
• xace
• xace_celegans

21
Phylogenics

• Clustal W
• Clustal X
• MOLPHY
• MrBayes
• PHYLIP

• RRTree
• T-Coffee
• TREE-PUZZLE
• TreeViewX

22
Structure Prediction

• EMBOSS
• MEME
• Modeller
• Mzef
• PHI-BLAST

23
Molecular Modelling

• Modeller
• homology modeling an alignment of a sequence to
  be modeled with known related structures
• Rasmol
• a molecular graphics program intended for 3D
  visualisation of proteins and nucleic acids
• Raster3D (publishing images)
• X3DNA
• analyzing and rebuilding 3D structures

24
Microarrays

• Dapple
• a program for quantitating spots on a two-colour
  DNA microarray image..
• OligoArray
• a program that computes gene specific
  oligonucleotides that are free of secondary
  structure for genome-scale oligonucleotide
  microarray construction.

25
Packages, Useful Scripts/Source Code, Graphics,
PERL

• BioPERL
• BioJava
• boxshade
• mvscf
• seg
• Split_fasta

• povRay
• Raster3D
• MOLPHY

26
Why We Need Supercomputers

• Some commercial packages run on supercomputers
• Accelrys modeling and simulation
• Materials Studio
• Cerius2 (SGI Unix only)
• Homology modeling to catalyst design
• Insight II (SGI Unix only)
• 3D graphical environment for physics based
  molecular modeling
• Catalyst (high end Unix servers)
• database management valuable in drug discovery
  research
• QUANTA (high end Unix servers)
• crystallographic 2D/3D protein structure solution
• Discovery Studio

27
Why We Need Supercomputers

• Supercomputer advantages
• Multiple processors
• Large shared memory
• Handle very large files
• Large/fast RAID arrays
• Terabyte tape backup systems
• Power backup systems
• High performance networks

28
Why We Need Supercomputers

• Common bioinformatics requirements
• Computationally intensive tasks
• Large memory models
• Intensive/complex database searches
• Large experimental database sets
• Large derived database sets
• Large persistent intermediate data structures
• Teamwork data sharing and visualization

29
Why We Need Supercomputers

• Network requirements
• Driving gigE/10gigE NICs
• Moving large files/data sets rapidly
• Visualization streams/Access GRID
• Coordinating Cluster/GRID computing
• Dynamic provisioning of light paths

30
Why We Need Supercomputers
31
Why We Need Supercomputers

xxxxxxxxxxxxxxxxxxxxxxx
32
Software Development Issues

• Collaboration contexts/barriers
• Team work collaboration spaces
• Standards development DTDs
• Integration issues
• experimental data to homology to 3D model
• platform issues
• network issues 9k MTU - jumbo frames
• Licensing issues public vs. private

33
Future Challenges

• Creating developer infrastructure for building up
  structural models from component parts
• components from macromolecule libraries ported to
  object models
• Understanding the design principles of systems of
  macromolecules and harnessing them to create new
  functions
• specialized molecular machines

34
Future Challenges

• Learning to design drugs efficiently and cost
  effectively based on knowledge of the target
• target generation automation
• validation automation
• Development of enhanced simulation models that
  give insight into context based function from
  knowledge of structure
• possible use of artificial intelligence to limit
  scope of search

35
How Tools might be used for Industry Biotech
Projects
36
Summary

• Bioinformatics
• well positioned to assist with application
  development
• exploring novel bioinformatics software
  development
• proceeding with supporting access GRID and
  optical switching technology

37
Questions/Comments ? -)