Bioinformatics: an overview

1
Bioinformatics an overview

• Ming-Jing Hwang (???)
• Institute of Biomedical Sciences
• Academia Sinica

http//gln.ibms.sinica.edu.tw/
2
The human genome project
Year 2001
3
Promises

• More will happen in biology in the next 10
• years than in the past 50
• (Craig Venter, Celera Genomics).
• We should be able to uncover the major
• hereditary contributions to common
• illnesses like diabetes and mental illness,
  
• probably in the next three to five years
• (Francis Collins, head of HGP).

4
Genetics Genomics From DNA to population
Source gsk
5
What makes us human ?

• The difference between you chimp is 1.24
• The difference between you and Maggie is 0.1

6
Hunting for disease genes
Source gsk
7
Genes and Diseases
Penetrance the likelihood that a person carrying
a particular mutant gene will have an altered
phenotype
Source gsk
8
phenotype and genotype

• Many different genotypes can have same phenotype
• Many genotypes do not change the phenotype
• One phenotype could be due to many different
  genotypes
• -- statistical genetics

9
The common variant common disease (CV-CD)
hypothesis
It is believed that most polygenic contributions
to disease susceptibility will arise from
variants that are relatively common in the
susceptible population.
10
Genetic variations

• SNP constitute 90 of human genetic variations
• Other forms of variations include insertion,
  deletion, and differences in the copy number of
  tandem repeats or large genomic segments, etc.

11
Three phases of human genome sequencing

• The genome map (draft in 2001, finished in
  2003)
• The SNP map (TSC, 2001)
• The haplotype map (HapMap, 2005)

12
Source gsk
13
Source gsk
14
pharmacogenomics
8/2 (?) 1000pm on PTS (CH13)
15
(Nature, 2004)
(PNAS, 2005)
16
Common SNPs
Kruglyak Nickerson, 2001
17
dbSNP summary (NCBI build 124)
July, 2005
18
Haplotype structure of the human genome
Goldstein, 2001
19
(No Transcript)
20
Rationale of HapMap
In a given population, 55 percent of people may
have one version of a haplotype, 30 percent may
have another, 8 percent may have a third, and the
rest may have a variety of less common
haplotypes. The International HapMap Project is
identifying these common haplotypes in four
populations from different parts of the world. It
also is identifying "tag" SNPs that uniquely
identify these haplotypes. By testing an
individual's tag SNPs (a process known as
genotyping), researchers will be able to identify
the collection of haplotypes in a person's DNA.
The number of tag SNPs that contain most of the
information about the patterns of genetic
variation is estimated to be about 300,000 to
600,000, which is far fewer than the 10 million
common SNPs.
21
Beyond genome
22
Chemical genomics
23
(No Transcript)
24
Bioinformatics has many sub-disciplines

• Genome Informatics (DNA sequence)
• Transcriptome Informatics (expression)
• Proteome informatics (ID, post-transl. mod.)
• Protein Informatics (protein struct./funct.)
• Evolutionary Informatics
• Biomedical Informatics (human disease)

25
Briefings in bioinformatics (Mar 2005)

• The many faces of sequence alignment (Altman )
• Bioinformatics analysis of alternative splicing
  (Lee Wang)
• Putting microarray in a context Integrated
  analysis of diverse biological data (Troyanskaya)
• Bioinformatics approaches and resources for
  single nucleotide polymorphism functional
  analysis (Mooney)
• A survey of current work in biomedical text
  mining (Cohen Hersh)
• Current efforts in the analysis of RNAi and RNAi
  target genes (Bengert Dandekar)

26
Sequence alignment

• The problem is still not solved.
• Sequence alignment methodology and tool
  development continue to grow, indicating that the
  alignment problem is still not solved.
• How can that be, after nearly forty years of
  research and literally hundreds of available
  tools?
• Why should alignments remain an open problem?
• It is not a single problem but rather a
  collection of many quite diverse questions that
  all have in common the search for sequence
  similarity
• The exponential expansion of biological sequence
  databases faster than Moores law

Batzoglou, 2005
27
Sequence alignment challenges

• Sensitivity and specificity
• Speed
• Evaluation
• Low similarity
• Rearrangements
• Orthology detection
• Multiple (genome) alignments

28
Evolution of functional important regions over
time
Miller et al., 2005
29
Evolutionary Informatics/ Comparative genomics
(Ureta-Vidal, Ettwiller, Birney, 2003)
30
Schema of genome alignment
(2003)
31
Genome alignment recent reviews

• An Applications-Focused Review of Comparative
  Genomics Tools Capabilities, Limitations and
  Future Challenges (Chain et al. Briefings in
  bioinformatics, 2003)
• The many faces of sequence alignment (Batzoglou,
  Briefings in bioinformatics, 2005)
• Comparative genomics (Miller et al. Annu. Rev.
  Genomics Hum. Genet. 2004)

32
RNAi post-translational gene regulaion

• Computational identification of miRNAs
• Computational prediction of miRNA targets
• miRNA data resources

33
Transcriptomics tools for understanding the body
plan
34
Microarray Integrated analysis
Troyanskaya, 2005
35
Proteomics
Initial goal identification of all proteins
expressed by a cell or tissue
36
From 1D to 3D The Holy Grail of Structural
Bioinformatics
MADWVTGKVTKVQNWTDALFSLTVHAPVLPFTAGQFTKLGLEIDGERVQR
AYSYVNSPDNPDLEFYLVTVPDGKLSPRLAALKPGDEVQVVSEAAGFFVL
DEVPHCETLWMLATGTAIGPYLSILRLGKDLDRFKNLVLVHAARYAADLS
YLPLMQELEKRYEGKLRIQTVVSRETAAGSLTGRIPALIESGELESTIGL
PMNKETSHVMLCGNPQMVRDTQQLLKETRQMTKHLRRRPGHMTAEHYW
37
(No Transcript)
38
Structural Bioinformatics Sequence/Structure
Relationship
Percent Identity
100 90 80 70 60 50 40 30 20 10 0
All possible sequences of amino acids
Protein structures observed in nature
Twilight zone
Midnight zone
Protein sequences observed in nature
39
Structure Prediction Methods
Homology modeling
Fold recognition
ab initio
0 10 20 30 40 50 60
70 80 90 100
sequence identity
40
CASP Experiments
41
Some CASP4 successes
Bakers group
42
3D to 1D?
Science 2003
43
A computer-designed protein (93 aa) with 1.2 A
resolution
44
Sequence/Structure Gap
Sequence
Structure
45
Structural Genomics solving fold representatives
Baker Sali, 2001
46
Structural Genomics overview

• When 1997 by Barry Honig, Wayne Henderickson and
  colleagues in a DOEs Advanced Photon Source
  (APS) proposal
• Goals 10,000 structures (100-200 str/center/yr)
  each representing a protein family in 5(10)
  years
• Enabling factors genome sequences, technology
  advancement (synchrotron MAD, etc.)
• Cost reducing current US200,000/str to
  10,000/str (est. 1.5-5 billion US)
• Players academic industry

47
Flowchart of a SG project
Burley etal., 1999
48
PSI phase I (pilot) centers

• Berkeley Structural Genomics Center focused on
  two bacterial species with extremely small
  genomes to study proteins essential for
  independent life. Principal investigator
  Sung-Hou Kim, Lawrence Berkeley National
  Laboratory
• Center for Eukaryotic Structural Genomics, based
  in Wisconsin, focused on protein production,
  characterization, and structure determination
  from Arabidopsis thaliana, a plant that is
  frequently used in laboratory research and that
  has many genes in common with humans and animals.
  Principal investigator John Markley, University
  of Wisconsin, Madison
• Joint Center for Structural Genomics, based in
  California, focused on novel structures from
  thermophilic microorganisms and on human proteins
  thought to be involved in cell signaling.
  Principal investigator Ian Wilson, The Scripps
  Research Institute
• Midwest Center for Structural Genomics, based in
  Illinois, selected bacterial targets related to
  disease and proteins from all three kingdoms of
  life. The emphasis was on previously unknown
  folds and on proteins from disease-causing
  organisms. Principal investigator Andrzej
  Joachimiak, Argonne National Laboratory
• New York Structural Genomics Research Consortium
  solved protein structures for disease-related
  proteins from eukaryotes and bacteria. Principal
  investigator Stephen K. Burley, Structural
  GenomiX, Inc.
• Northeast Structural Genomics Consortium, based
  in New Jersey, focused on target proteins from
  various model organisms, including the fruit fly,
  yeast, and roundworm. It used both X-ray
  crystallography and NMR spectroscopy. Principal
  investigator Gaetano Montelione, Rutgers
  University
• The Southeast Collaboratory for Structural
  Genomics, based in Georgia, determined structures
  from the prokaryotic model organism, Pyrococcus
  furiosus, and the eukaryotic model organism C.
  elegans, as well as some human proteins.
  Principal investigator Bi-Cheng Wang, University
  of Georgia
• Structural Genomics of Pathogenic Protozoa
  Consortium, based in Washington, solved protein
  structures from organisms known as protozoans,
  many species of which cause deadly diseases such
  as sleeping sickness, malaria, and Chagas'
  disease. Principal investigator Wim G. J. Hol,
  University of Washington
• TB Structural Genomics Consortium, based in New
  Mexico, analyzed protein structures from
  Mycobacterium tuberculosis. Principal
  investigator Thomas Terwilliger, Los Alamos
  National Laboratory

49
PSI Pilot Phase Facts at a Glance

• Goal To develop new approaches and tools needed
  to streamline and automate the steps of protein
  structure determination, and to incorporate those
  methods into high-throughput pipelines that use
  DNA sequence information to generate
  three-dimensional protein structure models
• Project period September 2000 to June 2005
• Funding 270 million (funded largely by the
  National Institute of General Medical Sciences,
  with additional support from the National
  Institute of Allergy and Infectious Diseases)
• Number of Centers 9 (6 survived to phase II)
• Solved protein structures More than 1,100
• Unique structures solved (structures sharing less
  than 30 percent of their sequence with other
  known proteins) More than 700

50
PDB content growth (May 2005)
51
Many bottlenecks remain target tracking by PDB
(Sep 2002)
52
Current (phase II) PSI centers
53
Hybrid approach for solving macromolecular
complex structures
54
Protein network an integrated approach
Aloy et al, 2004
55
Bioinformatics and Drug Design
Scientific America 2000
56
Yeast protein interaction network
Nat Rev Genet. 2004
57
Network parameters

• Degree (connectivity) k
• Degree distribution P(k), probability that a
  selected node has exactly k links.
• Scale-free network degree distribution
  approximates a power law, P(k) k-? (? degree
  exponent)

Log(P(k))
Log(k)
Barabasi Oltvai, Nat Rev Genet. 2004
58
Network models
Barabasi Olvtai, Nat Rev Genet. 2004
59
Scale-free networks
P(k) k-?, (?in in-degree ?out out-degree
exponent)
Albert Barabasi, Reviews of Modern Physics, 2002
60
Challenges in network biology

• Network databases
• Information integration
• Organization characteristics and principles
• Design rules
• Evolution mechanisms
• Validation

61
Neuroinformatics neuroscience bioinformatics
The human brain project -UC Davis http//nir.cs.uc
davis.edu/index.jsp
http//ncmir.ucsd.edu/NCDB/
62
Bioinformatics Journals

• Bioinformatics
• Nucleic Acids Research
• BMC Bioinformatics
• Briefings in Bioinformatics
• Proteins
• J. Mol. Biol.

• PNAS
• PLoS computational biology
• Genome Research

63
Scope of bioinformatics

• Genome analysis
• Sequence analysis
• Phylogenetics
• Structural bioinformatics
• Gene expression
• Genetic and population analysis
• Systems biology
• Data and text mining
• Databases and ontologies

64
Sample articles of a recent issue

• Exondomain correlation and its corollaries
• Functional annotation from predicted protein
  interaction networks
• HYPROSP II-A knowledge-based hybrid method for
  protein secondary structure prediction based on
  local prediction confidence
• Comparative interactomics analysis of protein
  family interaction networks using PSIMAP (protein
  structural interactome map)
• Semi-supervised protein classification using
  cluster kernels
• A new progressive-iterative algorithm for
  multiple structure alignment
• Practical FDR-based sample size calculations in
  microarray experiments
• Mining genetic epidemiology data with Bayesian
  networks I Bayesian networks and example
  application (plasma apoE levels)
• Inferring proteinprotein interactions through
  high-throughput interaction data from diverse
  organisms
• A latent variable model for chemogenomic
  profiling

65

• NAR Database issue (Jan. 2005)

Categories
1. Nucleotide Sequence Databases 53
2. RNA Sequence Databases 34
3. Protein Sequence Databases 105
4. Structure Databases 64
5. Genomic Databases (non-human) 134
6. Metabolic Enzyme Pathways Signals Pathways 36
7. Human Other Vertebrate Genomes 64
8. Human Genes Diseases 69
9. Microarray Data Other Gene Expression Databases 42
10. Proteomics Resources 7
11. Other Molecular Biology Databases 17
12. Organelle Databases 18
13. Plant Databases 48
14. Immunological Databases 20
Total 711
http//nar.oupjournals.org/cgi/content/full/33/sup
pl_1/D5/TBL1
66
NAR Web Server Issue (July 2005)
Year of publication
2004 129 (137)
2005 166
Total 295
67
Computer Related (2) Bio- Programming Tools
(1) Statistics (1)DNA (57) Annotations
(9) Gene Prediction (4) Mapping and Assembly
(1) Phylogeny Reconstruction (4) Sequence
Feature Detection (16) Sequence Polymorphisms
(8) Sequence Retrieval and Submission (3) Tools
For the Bench (12)Education (1) Directories and
Portals (1)Expression (48) cDNA, EST, SAGE
(8) Gene Regulation (22) Microarrays
(16) Splicing (2) Human Genome (13) Annotations
(3) Health and Disease (3) Other Resources
(2) Sequence Polymorphisms (5)Model Organisms
(9) Microbes (4) Mouse and Rat (2) Plants
(1) Yeast (2)Other Molecules (2) Carbohydrates
(2)
Protein (131) 2-D Structure Prediction (10) 3-D
Structure Prediction, Comparison (34) 3-D
Structure Retrieval, Viewing (6) Biochemical
Features (8) Domains and Motifs (25) Function
(10) Interactions, Pathways, Enzymes
(13) Localization and Targeting (7) Phylogeny
Reconstruction (5) Proteomics (2) Sequence
Features (6) Sequence Retrieval (5)RNA
(15) Functional RNAs (5) Motifs (3) Sequence
Retrieval (2) Structure Prediction,
Visualization, and Design (5)Sequence Comparison
(29) Alignment Editing and Visualization
(2) Analysis of Aligned Sequences
(12) Comparative Genomics (7) Multiple Sequence
Alignments (2) Pairwise Sequence Alignments
(2) Similarity Searching (4) Literature
(5) Search Tools (3) Text Mining (2)
http//bioinformatics.ubc.ca/resources/links_direc
tory/narweb2005/
68
NAR database issue (Jan. 2005)
Categories URL not available /not working Not recommended Recommended
1. Nucleotide Sequence Databases 53 3 41 9
2. RNA Sequence Databases 34 4 23 7
3. Protein Sequence Databases 104 4 66 34
4. Structure Databases 64 7 9 48
5. Genomic Databases (non-human) 134 11 105 17
6. Metabolic Enzyme Pathways Signals Pathways 36 3 20 13
7. Human Other Vertebrate Genomes 62 2 42 18
8. Human Genes Diseases 69 4 54 11
9. Microarray Data Other Gene Expression Databases 42 4 31 7
10. Proteomics Resources 7 0 5 2
11. Other Molecular Biology Databases 17 1 10 6
12. Organelle Databases 18 1 13 4
13. Plant Databases 48 11 36 1
14. Immunological Databases 20 0 17 3
Total 708 55 474 179
138 can be downloaded
69
An example of our curation
70
Two keys in bioinformatics research

• Solve a significant biological question
• Develop a must-use application tool