Analysis Environments For Functional Genomics

1
Analysis Environments For Functional Genomics
Bruce R. Schatz CANIS LaboratoryInstitute for
Genomic BiologyUniversity of Illinois at
Urbana-Champaign schatz_at_uiuc.edu ,
www.canis.uiuc.edu
Bioinformatics Seminar Department of Computer
Science, UIUC February 25, 2005
2
What are Analysis Environments

• Functional Analysis
• Find the underlying Mechanisms
• Of Genes, Behaviors, Diseases
• Comparative Analysis
• Top-down data mining (vs Bottom-up)
• Multiple Sources especially literature

3
Building Analysis Environments

• Manual by Humans
• Interaction user navigation
• Classification collection indexing
• Automatic by Computers
• Federation search bridges
• Integration results links

4
Trends in Analysis Environments

• Central versus Distributed Viewpoints
• The 90s Pre-Genome
• Entrez (NIH NCBI) versus
• WCS (NSF Arizona)
• The 00s Post-Genome
• GO (NIH curators) versus
• BeeSpace (NSF Illinois)

5
Pre-Genome Environments

• Focused on Syntax pre-Web
• WCS (Worm Community System)
• Search words across sources
• Follow links across sources
• Words automatic, Links manual
• Towards Uniform Searching

6
Post-Genome Environments

• Focused on Semantics post-Web
• BeeSpace (Honey Bee Inter Space)
• Navigate concepts across sources
• Integrate data across sources
• Concepts automatic, Links automatic
• Towards Question Answering

7
Paradigm Shift

• Towards Dry-Lab Biology, Walter Gilbert (Jan
  1991)
• The new paradigm, now emerging, is that all the
  'genes' will be known (in the sense of being
  resident in databases available electronically),
  and that the starting point of a biological
  investigation will be theoretical. An
  individual scientist will begin with a
  theoretical conjecture, only then turning to
  experiment to follow or test that hypothesis.
  ...
• To use this flood of knowledge the total
  sequence of the human and model organisms, which
  will pour across the computer networks of the
  world, biologists not only must become
  computer-literate, but also change their approach
  to the problem of understanding life. ...
• The Coming of Informational Science
• Correlation of Information across Sources

8
NCBI Entrez
9
Community Systems
results
data
(database management)
(electronic mail)
knowledge
(hypertext annotations)
literature
news
(information retrieval)
(bulletin boards)
Formal
Informal
browse and share all the knowledge of a community
10
Worm Community System

• WCS Information
• Literature BIOSIS, MEDLINE, newsletters,
  meetings
• Data Genes, Maps, Sequences, strains, cells
• WCS Functionality
• Browsing search, navigation
• Filtering selection, analysis
• Sharing linking, publishing
• WCS 250 users at 50 labs across Internet (1991)

11
WCS Molecular
12
WCS Cellular
13
WCS Publishing
14
WCS Linking
15
WCS invokes gm
16
WCS vis-à-vis acedb
17
Towards the Interspace

• from Objects to Concepts
• from Syntax to Semantics
• Infrastructure is Interaction with Abstraction

Internet is packet transmission across
computers Interspace is concept navigation
across repositories
18
THE THIRD WAVE OF NET EVOLUTION
CONCEPTS
OBJECTS
PACKETS
19
LEVELS OF INDEXES
20
COMPUTING CONCEPTS
92 4,000 (molecular biology) 93 40,000
(molecular biology) 95 400,000 (electrical
engineering) 96 4,000,000 (engineering) 98
40,000,000 (medicine)
21
Simulating a New World

• Obtain discipline-scale collection
• MEDLINE from NLM, 10M bibliographic abstracts
• human classification Medical Subject Headings
• Partition discipline into Community Repositories
• 4 core terms per abstract for MeSH classification
• 32K nodes with core terms (classification tree)
• Community is all abstracts classified by core
  term
• 40M abstracts containing 280M concepts
• concept spaces took 2 days on NCSA Origin 2000
• Simulating World of Medical Communities
• 10K repositories with gt 1K abstracts (1K w/ gt
  10K)

22
Interspace Remote Access Client
23
Navigation in MEDSPACE

• For a patient with Rheumatoid Arthritis
• Find a drug that reduces the pain (analgesic)
• but does not cause stomach (gastrointestinal)
  bleeding

Choose Domain
24
Concept Search
25
Concept Navigation
26
Retrieve Document
27
Navigate Document
28
Retrieve Document
29
Informational Science

• Computational Science is widely accepted as
• The Third Branch of Science
• (beyond Experimental and Theoretical)
• Genes are Computed, Proteins are Computed,
• Sequence equivalences are Computed.
• Informational Science is coming to be accepted as
• The Fourth Branch of Science
• Based on Information Science technologies for
• Functional Analysis across Information Sources

30
Post-Genome Informatics I

• Comparative Analysis within the
• Dry Lab of Biological Knowledge
• Classical Organisms have Genetic Descriptions.
• There will be NO more classical organisms beyond
• Mice and Men, Worms and Flies, Yeasts and Weeds.
• Must use comparative genomics on classical
  organisms
• Via sequence homologies and literature analysis.

31
Post-Genome Informatics II

• Functional Analysis within the
• Dry Lab of Biological Knowledge
• Automatic annotation of genes to standard
  classifications, e.g. Gene Ontology via homology
  on computed protein sequences.
• Automatic analysis of functions to scientific
  literature, e.g. concept spaces via text
  extractions. Thus must use functions in
  literature descriptions.

32
Informatics From Bases to Spaces

• data Bases support genome data
• e.g. FlyBase has sequences and maps
• Genes annotated by GO and linked to literature
• e.g. BeeBase has computed annotations
• Protein homologies for similar Genes via GO
• information Spaces support biomedical literature
• e.g. BeeSpace uses automatically generated
• conceptual relationships to navigate functions

33
Gene Ontology
34
Gene Ontology

• Gene Symbol Data Source Full Name
• Calca MGI calcitonin-related polypeptide
• Cat-1 Wormbase None
• Cat-2 Wormbase None
• CCKR-Human UniProt Cholecystokinin receptor
• CRF2-Rat UniProt Corticotropin releasing factor
• Crhr2 RGD corticotrophin relse hormone
• Egl-10 Wormbase None
• Egl-30 Wormbase None
• Feh-1 Wormbase None
• For FlyBase None

35
Conceptual Navigation in BeeSpace
36
BeeSpace Analysis Environment

• Build Concept Space of Biomedical Literature for
  Functional Analysis of Bee Genes
• -Partition Literature into Community Collections
• -Extract and Index Concepts within Collections
• -Navigate Concepts within Documents
• -Follow Links from Documents into Databases
• Locate Candidate Genes in Related Literatures
  then follow links into Genome Databases

37
Question Answering
38
Functional Phrases

• ltgenegt encodes ltchemicalgt
• Sokolowski and colleagues demonstrated in
  Drosophila melanogaster that the foraging gene
  (for) encodes a cGMP dependent protein kinase
  (PKG).
• The dg2 gene encodes a cyclic guanosine
  monophosphate (cGMP)- dependent protein kinase
  (PKG).
• ltchemicalgt affects/causes ltbehaviorgt
• Thus, PKG levels affected food-search behavior.
• cGMP treatment elevated PKG activity and caused
  foraging behavior.
• ltgenegt regulates ltbehaviorgt
• Amfor, an ortholog of the Drosophila for gene, is
  involved in the regulation of age at onset of
  foraging in honey bees.
• This idea is supported by results for malvolio
  (mvl), which encodes a manganese transporter and
  is involved in regulating Drosophila feeding and
  age at onset of foraging in honey bees.

39
BeeSpace Software Implementation

• Natural Language Processing
• Identify noun and verb phrases
• Recognize biological entities
• Compute biological relations
• Statistical Information Retrieval
• Compute statistical contexts
• Support conceptual navigation

40
Data Integration (FlyBase Gene)

• D. melanogaster gene foraging , abbreviated as
  for , is reported here . It has also been known
  in FlyBase as BcDNAGM08338, CG10033 and
  l(2)06860. It encodes a product with
  cGMP-dependent protein kinase activity
  (EC2.7.1.-) involved in protein amino acid
  phosphorylation which is a component of the
  cellular_component unknown . It has been
  sequenced and its amino acid sequence contains an
  eukaryotic protein kinase , a protein kinase
  C-terminal domain , a tyrosine kinase catalytic
  domain , a serine/Threonine protein kinase family
  active site , a cAMP-dependent protein kinase and
  a cGMP-dependent protein kinase . It has been
  mapped by recombination to 2-10 and cytologically
  to 24A2--4 . It interacts genetically with Csr .
  There are 27 recorded alleles 1 in vitro
  construct (not available from the public stock
  centers), 25 classical mutants ( 3 available from
  the public stock centers) and 1 wild-type.
  Mutations have been isolated which affect the
  larval nerve terminal and are behavioral, pupal
  recessive lethal, hyperactive, larval
  neurophysiology defective and larval neuroanatomy
  defective. for is discussed in 80 references
  (excluding sequence accessions), dated between
  1988 and 2003. These include at least 6 studies
  of mutant phenotypes , 2 studies of wild-type
  function , 3 studies of natural polymorphisms and
  7 molecular studies . Among findings on for
  function, for activity levels influence adult
  olfactory trap response to a food medium
  attractant. Among findings on for polymorphisms,
  the frequency of for R and for s strains in three
  natural populations are studied to determine the
  contribution of the local parasitoid community to
  the differences in for R and for s frequencies.

41
BeeSpace Information Sources

• Biomedical Literature
• Medline (medicine)
• Biosis (biology)
• Agricola, CAB Abstracts, Agris (agriculture)
• Model Organisms (heredity)
• -Gene Descriptions (FlyBase, WormBase)
• Natural Histories (environment)
• -BeeKeeping Books (Cornell Library, Harvard Press)

42
Medical Concept Spaces (1998)

• Medical Literature (Medline, 10M abstracts)
• Partition with Medical Subject Headings (MeSH)
• Community is all abstracts classified by core
  term
• 40M abstracts containing 280M concepts
• computation is 2 days on NCSA Origin 2000
• Simulating World of Medical Communities
• 10K repositories with gt 1K abstracts
• (1K with gt 10K)

43
Biological Concept Spaces (2005)

• Compute concept spaces for All of Biology
• BioSpace across entire biomedical literature
• 50M abstracts across 50K repositories
• Use Gene Ontology to partition literature into
• biological communities for functional analysis
• GO same scale as MeSH but adequate coverage?
• GO light on social behavior (biological process)

44
Paradigm Shift

• Dissecting Human Disease, Victor McKusick (Feb
  2001)
• Structural genomics Functional genomics
• Genomics Proteomics
• Map-based gene discovery Sequence-based gene
  discovery
• Monogenic disorders Multifactorial disorders
• Specific DNA diagnosis Monitoring susceptibility
• Analysis of one gene Analysis of multi-gene
  pathways
• Gene action Gene regulation
• Etiology (mutation) Pathogenesis (mechanism)
• One species Several species

45
Needles and Haystacks

• Genes
• Honey Bees have 13K genes
• Perhaps 100 have known functions
• Paths
• Perhaps 30K protein families exist
• KEGG has 200 known pathways
• Statistical Clustering for Interactive Discovery
• Across Two Orders of Magnitude!

46
Concept Switching

• In the Interspace
• each Community maintains its own repository
• Switching is navigating Across repositories
• use your specialty vocabulary to search another
  specialty

47
CONCEPT SWITCHING

• Concept versus Term
• set of semantically equivalent terms
• Concept switching
• region to region (set to set) match

48
Biomedical Session
49
Categories and Concepts
50
Concept Switching
51
Document Retrieval
52
Future Technologies

• Concept Switching
• Spreading activation, type tagging
• Dynamic Indexing
• On-the-fly collections, during session
• Path Matching
• Aggregating indexes, many repositories

53
THE NET OF THE 21st CENTURY

• Beyond Objects to Concepts
• Beyond Search to Analysis
• Problem Solving via Cross-Correlating Multimedia
  Information across the Net
• Every community has its own special library
• Every community does semantic indexing
• The Interspace approximates Cyberspace

54
Interactive Functional Analysis

• BeeSpace will enable users to navigate a uniform
  space of diverse databases and literature sources
  for hypothesis development and testing, with a
  software system beyond a searchable database,
  using literature analyses to discover functional
  relationships between genes and behavior.
• Genes to Behaviors
• Behaviors to Genes
• Concepts to Concepts
• Clusters to Clusters
• Navigation across Sources

55
XSpace Information Sources

• Organize Genome Databases (XBase)
• Compute Gene Descriptions from Model Organisms
• Partition Scientific Literature for Organism X
• Compute XSpace using Semantic Indexing
• Boost the Functional Analysis from Special
  Sources
• Collecting Useful Data about Natural Histories
• e.g. CowSpace Leverage in AIPL Databases

56
Towards the Interspace

• The Analysis Environment technology is
  GENERAL! BirdSpace? BeeSpace?
• PigSpace? CowSpace?
• BehaviorSpace? BrainSpace?
• BioSpace
• Interspace