cancer cell signaling with PATIKA

Document Sample
cancer cell signaling with PATIKA Powered By Docstoc
					Cancer Cell Signaling Database
        with PATIKA




          Asst. Prof. Rengül Çetin-Atalay,
         Bilkent University, Ankara, Turkey
   Department of Molecular Biology and Genetics,
              & Bioinformatics Center
                                                   1/66
                Outline

•   Why pathway informatics
•   Related work
•   PATIKA ontology
•   PATIKA software
•   i-Cancer



                              2/66
                                Cancer




                                                   D. Hanahan and R. A. Weinberg, Cell, 2000


Carcinogenesis involves aberrations in highly complex interconnected signalling
   systems of cell.

Any confusion in taking decisions during signalling in a time- and cell-dependent
   manner may result in uncontrolled cell behaviour (enhanced proliferation,
   resistance to cell death, defect in sensing DNA damage …).
                                                                                               3/66
 Hepatocellular carcinoma
Genes that are involved in Hepatocellular carcinona.
                                           IGF2R
                               TGFBR2              p16
                                Smad2       p53        pRb
                     c-Met                                      TGF-
                                  Smad4        Cyclin D
                   -catenin                                   Integrins
                                         Resistance
             E-cadherin                                                c-Met
                                         to Growth

                                 tas &
                                          Inhibition                     APC

                                    is
            Integrins     Me sion




                                                       Gro
                                                       Gro
                                                       Se h
                                                       Sel th
             MMP                                                           Axin
                             tas
                               a




                                                          f
                                                          f
                                                           w
                           Inv


            TIMP                                                         -catenin




                                                           pop nce
                         An
                         An




                                                                           IGF2




                                                                  is
                                                               tos
            VEGF
                           Ne enesi
                           Ne enes
                           go
                           giio




                                                       to A sista
                                                                         IGF2R
                               o
                               o-
                                g
                                g




             Integrins                                    Re           TGFBR2
                                         Immortality
                                   s
                                   s




                                                                p53     PTEN
               Extracellular
                 Proteases                                        FasL
                                         Telomerase               Fas
                                    p53                pRb
                                            p16
                                                                                     4/66
                       High Throughput data


                                    Yeast Two-Hybrid
                Tissuearray

                                                              Proteomics
Microarray
                    Expression, Interaction and Post
                     transnational modification data




                                 Data
Protein Databases             Acquisition              Scientific Literature
                              Integration
                               Analysis
                                                                           5/66
    Potential Products of Data Acquisition
           Integration and Analysis
• Generate metabolic, signaling, and other
  pathways
• Attribute new functions to known gene products
• Find new genes and assign them functions
• Predict novel protein interaction pathways
• link genotype to phenotype
• Compare cross-species
• Design drugs for target proteins
• Design new clinical tests
                                                   6/66
Why large scale expression analysis




                                      7/66
       Millions of Interaction

The complexity of the system is fairly high.

Starting with the assumption that there are
 approximately 25 000-28 000 genes in the
 human genome one easily realizes that
 number of entities one need to consider for
 analyzing a single cell is way beyond a
 million.

                                               8/66
Complexity




             9/66
Complexity




             10/66
                  Outline

•   Why pathway informatics
•   Related work
•   PATIKA ontology
•   PATIKA software
•   i-Cancer



                              11/66
                              Approaches
There are already significant amount of work on pathway modeling and pathway databases

Still Image Databases                            Interaction Databases
Easy data generation                             Aim to cope with rapidly emerging
Conventional                                     protein-protein and protein-NA
                                                 interaction data
Knowledge can not be integrated
                                                 Mainly deal with intermolecular
automatically
                                                 interactions
No meaningful querying, simulation
                                                 Low efficiency in representing
and prediction is possible                       complex cellular processes
                  Metabolic Pathway Databases
                  More elucidated compared to signaling
                  pathways
                  Contain structured, extensive data on
                  metabolic events
                  Visualization is often provided as a set of
                  still images or dynamic graphs                                     12/66
   New Definitions: Parameters
          for an Ontology
–Coverage
–Comprehensibility &Visualization.
–Querying & Analysis.
–Manipulation & Integration

Signaling pathway databases
                                     13/66
Databases of Protein Interactions
      General Pathway
      PATIKA - Pathway Analysis Tool for Integration and Knowledge Acquisition
      MINT - a database of Molecular INTeractions
      KEGG - Kyoto encyclopedia of genes and genomes
      Transpath - Biobase homepage
      BioCyc – Peter Karp
      WITT
      BindingDB - The Binding Database
      SPAD - signaling pathway
      PIM - Protein Interaction Map
      BioCarta
      PFBP - Protein Function and Biochemical Pathways Project
      PathSCOUT- Lion Biosciences
      CSNDB - Cell Signaling Networks Database
      BRITE - Biomolecular Relations in Information Transmission and Expression
      Expasy - Biochemical pathways
      EMP - The Enzymology Database
      FIMM - A Database of Functional Molecular Immunology
      Binary Interaction
      BIND - Biomolecular Interaction Network Database
      DIP - Database of Interacting Proteins
      DPInteract - DNA-protein interactions
      Interact - A Protein-Protein Interaction database


                                                                                  14/66
   BioCarta




http://www.biocarta.com
                          15/66
           BIND: Biomolecular Interaction Network Database


mcbrowse
ibrowse




       15504 Interactions
              and
          8 Pathways
                                       http://www.bind.ca/
                                                             16/66
                               KEGG
Kyoto encyclopedia of genes and genomes
     Database                     Content                                Source
                Protein interaction networks, such as
                                                             Manually entered from
PATHWAY         pathways and complexes, that are
                                                             published materials
                responsible for various cellular processes
                GENES: Gene catalogs for completely
                sequenced genomes and some partial           Generated from GenBank,
                genomes                                      RefSeq, and other public
                                                             resources with reannotation by
GENES and       GENOME: Genome maps and organism             KEGG
associated      information
databases       EXPRESSION: Microarray gene                  Microarray data obtained by the
                expression profiles                          Japanese groups
                BRITE: Protein-protein interactions and      Two-hybrid and other data that
                relations                                    are publicly available
                                                           Computationally derived from
                Sequence similarities as well as
                                                           GENES by pairwise genome
SSDB            ortholog/paralog relations among all genes
                                                           comparisons of all protein-
                in all organisms
                                                           coding genes
                                                             Manually entered using SSDB
KO              KEGG Orthology (KO) grouping
                                                             classification
                COMPOUND: Chemical compound
                structures                                   Manually entered from
                REACTION: Chemical reactions                 published materials
LIGAND
                GLYCAN: Carbohydrate structures
                ENZYME: Enzyme and enzymatic                 Generated from
                reactions                                    IUBMB/IUPAC nomenclature

                                                                                               17/66
KEGG:Regulatory pathways




                           18/66
         TRANSPATH ®                 Professional




12.262 Molecule
15.346 Reaction
                  Biobase http://www.biobase.de
                       Commercial                   19/66
Path-SCOUT and p-SCOUT




   LION Bioscience, Commercial
                                 20/66
                  BioCyc




Literature-derived Pathway/Genome Databases by Peter
Karp
EcoCyc-- Escherichia coli
MetaCyc-- Metabolic pathways and enzymes from 150 species   21/66
      Collaborative Affords for Pathway Informatics

BoiPAX:Biological Pathways              BioPathways Consortium Working
Exchange                                Groups
The goal of the BioPAX group is to      The goal of the BioPathways
develop a common exchange format        Consortium is to catalyze the
for biological pathways data.           emergence and development of
                                        computational pathways biology, by
Databases                               building up a strong and coherent
Representatives from a number of        scientific community, sharing
existing databases have been involved   knowledge, facilitating collaborations,
in the BioPAX effort, including:        and fostering the development of
•BioCyc                                 methods and tools of wide interest to
•BIND                                   the community.
•WIT
Group                                   Group
•Proteomics Standards Initiative        •Pathway Text-Mining, Geneways at
•Chemical Markup Language               Columbia Univeristy
•SBML                                   •Ontologies and Formalisms,PathOS
•CellML                                 •BioPAX

                                                                                  22/66
         PATIKA
Pathway Analysis Tool for Integration
    and Knowledge Acquisition




                                        23/66
                  Outline

•   Why pathway informatics
•   Related work
•   PATIKA ontology
•   PATIKA software
•   i-Cancer



                              24/66
                      AIM
PATIKA project aims to provide an integrated
 pathway editing environment with:
  – A well defined ontology,
  – A net based open knowledge base for integrating
    current knowledge on pathways,
  – Powerful querying and analysis options,
  – Dynamic visualization of any given subgraph of
    database.



                                                      25/66
 Ontology: Life Cycle of a Molecule

Either synthesized from its
   precursors or transported
   into the cell,
Then it is (optionally)
   modified or transported to
   a specific location in the
   cell,
Finally it is either degraded
   or transported out of the
   cell.


                                      26/66
              Modeling Life

In a cell, decision mechanisms are governed
   by molecules.
It is possible for a molecule to go through a
   certain subset of possible events at a certain
   condition and a totally different set of
   events at another condition.
In fact this is how a cell decides how to react
   to an input.

                                                    27/66
             PATIKA Ontology

In PATIKA ontology each of these changes that molecules go
   through is called transition,
A transition can be:
   – Chemical modification, complex formation, membrane transport,
     e.t.c.


Each variant of a certain molecule is called a state of the
  molecule.
A state can be;
   – A macromolecule (DNA, RNA, Protein), a small molecule (ions,
     drugs, hormones), a complex (ribosome), e.t.c.


                                                                     28/66
             PATIKA Ontology

Every state can be
associated with one or       t
more transitions.
A state is either
   – Substrate
   – Product
   – Effector
     (activator/inhibitor)



                                 29/66
                PATIKA Ontology
                                         Group Addition
                                         Group Removal
                                         Allosteric Change
             Chemical Modification
                                         Redox
             Replication

             Transcription               Alternative splicing


Transition   Translation
                                         Association         Multimerization
             Non-Covalent Modification
                                         Dissociation
             Membrane Transport

             Cleavage


                                                                        30/66
          PATIKA Ontology

States of common origin are grouped in
  Bioentities. For example;
  – States p53-phosphorylated, p53-native and p53-
    degraded can be grouped under the bioentity
    p53.
Bioentities are not actually drawn but act as
  data-holders


                                                     31/66
             PATIKA Ontology

A group of states
  that come together
  to perform certain
  functions, is called
  molecular complex.

Interactions among the states
   of complexes are also
   defined as bind relations.


                                32/66
              PATIKA Ontology

In a cell, molecules are
confined in
compartments, the sub-
cellular well defined
locations.




                                33/66
           PATIKA Ontology

Abstractions:
  Abstractions are compound graphs used for
  defining a certain subgraph of the big picture.
   – Regular abstractions
   – Homology Abstractions
   – Incomplete Abstractions



                                                    34/66
          PATIKA Ontology

Regular Abstraction:
  – Helps better handling of complex information.
  – Represents a certain abstraction of a biological
    phenomenon such as cell death,
  – Represented as connected sub graph of
    transitions and states,




                                                       35/66
           PATIKA Ontology
Abstractions help better handling of complex information.




                               Part of a pathway graph may be
                               “collapsed” to simplify a relatively
                               more complex pathway graph.

                                                                      36/66
              PATIKA Ontology
Incomplete information representing: Transition abstraction




       It is unknown whether S4 activates t1 or t2

                                                              37/66
            PATIKA Ontology
Incomplete information representing: State abstraction




        It is unknown whether S1 or S1' inhibits t2


                                                         38/66
         PATIKA Ontology

Homologies form another type of an abstraction.




                                                  39/66
               PATIKA Ontology


Canonical wnt pathway
represented by our
ontology




                                 40/66
Complexity of Molecular Networks




                                   41/66
          PATIKA USERS
                             Regular

Regular


                   PATIKA
                   Server

          Expert            Research



                                       42/66
                  Outline

•   Why pathway informatics
•   Related work
•   PATIKA ontology
•   PATIKA software
•   i-Cancer
•   Biospice


                              43/66
              PATIKA Architecture

   PATIKA              PATIKA
                                  RMI
    Editor             Managers

                                        PATIKA
                            RMI         Database
                HTTP
     Proxy             Servlet
                       Manager

Client Side                       Server Side
                                                   44/66
           Implementation

PATIKA is implemented in JAVA using a full
 object oriented approach.
TomSawyer™ Software‟s Graph Editor
 Toolkit for JAVA.
ObjectStore ™ database management system.




                                             45/66
Client Side-Detailed

                             PATIKA Editor
      User Wizard             PATIKA                   Simulator
      Manages users and       Core visual editor       Basic simulation
      user profiles                                    support

      Submission Wizard                                Layout Manager
      Submission Interface                             Manages
                                                       automated layout
                               Database-Editor
      Query Wizard             Converter               Inspector
      Querying interface       Converts database
                                                       Displays and
                               level objects to
                                                       manipulates an
                               editor level and vice
                                                       objects attributes
                               versa




                                     Proxy
                                                                            46/66
47/66
48/66
49/66
50/66
Layout before




                51/66
Layout after




               52/66
Server Side-Detailed

    Database                          PATIKA Database
    Manager
                                      Main Database that stores users,
    Performs database                 pathways and submissions.
    system specific
    tasks




   Query Manager           User Manager               Submission Manager
   Performs queries        Manages users              Manages submission
                           and user profiles          merges.




                           Servlet Manager
                        Manages incoming requests
                                                                           53/66
     Visual Query Tools:
Find States of Bioentity Dialog




                                  54/66
     Visual Query Tools:
Find Neighbour States Dialog




                               55/66
       Visual Query Tools:
Find Paths Between States Dialog




                                   56/66
        PATIKA QUERY TOOLS

PATIKA visual queri tools
 Lack of fexibility in implementing the overall
 system control structure
 Inability to handle complex queries
Requirements
 Find a path with complex queries
 have more complex pathways more than
 static visual query
 One output will be input of other basic query

                                                  57/66
           PATIKA QUERY TOOLS
             Query by Scripting
Script-based programming uses a direct representation of user's actions as
   a script.


Query by Scripting
   Script language Data structure and flow control structure
   Primitive pathway query operation definition (a Minimal set of operation
   on pathway ie shortest path, upregualtor, biologically meaningfull
   operations )
This generates PATIKA query by scripting



Jython is an implementation of the high-level, dynamic, object-oriented
     language
Python written in pure Java, and integrated with the Java platform.
It thus allows us to run python on any Java platform.
                                                                              58/66
Jython Query Dialog




                      59/66
                          Query by Scripting
Query 4: Find the states that are in 2 neighborhood of states which
  activates the transitions that state that is inhibited by „s2‟.
    sqc1.setName(‘s2’)
    set1 = sm.getStatesSatisfying(sqc1)
    set2 = sm.getInhibitedTransitionsOfStateSet(set1)
    set3 = sm.getActivatorStatesOfTransitionSet(set2)
    sm.getNeighbourhood(set3, 2, set4)
    finalResult=set4




                                                                      60/66
                      Primitif Operations
State-Oriented methods                              BioEntity-Oriented methods
- Set getStatesSatisfying(Set setOfObjects,         Set getBioEntitiesSatisfying (Set setOfObjects, Set
QueryConditions conditions)                         conditions)
- Set getStatesSatisfying(BioNodeQueryConditions    - Set
                                                    getBioEntitiesSatisfying(BioEntityQueryCondition
conditions)                                         s conditions)
- Set getStatesSatisfying(Set setOfObjects, Set     BioNode-Oriented methods
conditions)                                         - Set getBioNodesSatisfying(BioNo
 Interaction-Oriented methods                       deQueryConditions conditions)
- Set getStatesOfInteraction(DBInteraction          - Set getBioNodesSatisfying(Set setOfObjects, Set
anInteraction)                                      conditions)
- Set getSourceOfInteraction(DBInteraction          All Objects-Oriented methods
anInteraction)                                      - Set getAllBioNodes()
- Set getTargetOfInteractionSet(Set                 - Set getAllStates()
setOfInteractions)                                  - Set getAllInteractions()
 Transition-Oriented methods                         Other methods
- Set getSubstrateStatesOfTransition(DBTransition   - boolean isDownStream(DBBioNode sourceNode,
aTransition)                                        DBBioNode targetNode)
                                                    - boolean isUpRegulateAny(DBState sourceState,
- Set getProductStatesOfTransitionSet(Set           Set setOfStates)
setOfTransitions)
                                                    - Set getStatesUpRegulatingAllTargets(Set
- Set getProductStatesOfTransition(DBTransition     sourceSet, Set targetSet)
aTransition)
                                                                                                 61/66
         www.patika.org
PATIKA team




     Faculty Members                Graduate Students
     Atilla Gursoy PhD, CS          Emek Demir MS, MBG
     Mehmet Ozturk PhD, MBG         Ozgun Babur BS, MBG
     Rengul Cetin-Atalay PhD, MBG   Gurcan Gulesir BS, CS
     Ugur Dogrusoz PhD, CS          Gurkan Nisanci BS, CS
                                                            62/66
                  Outline

•   Why pathway informatics
•   Related work
•   PATIKA ontology
•   PATIKA software
•   i-Cancer



                              63/66
     Applications of iCANCER

Rapid Knowledge Acquisition
   – Provides a single source for cancer pathway information
   – Integrates our current knowledge on molecular biology of cancer
High throughput data analysis
   – Provides mechanistic explanations to array data
Drug Design
   – Rapid evaluation of drug candidates
   – Prediction of side-effects
   – Analysis of drug combinations for chemotherapy
Disease Gene Identification

                                                                       64/66
            iCancer Team
Dr. Mehmet Ozturk, Bilkent U. MBG
Dr. Ozlen Konu, Bilkent U. MBG
Dr. Rengul Atalay, Bilkent U. MBG
Dr. Veysi Isler, Mobilsoft
Halil Acu, MSc, Mobilsoft
Emin Deniz Ozkan, Bilkent U. MBG
Ahmet Rasit Ozturk, Bilkent U. MBG

                                     65/66
Bilkent University




                     66/66
t University




                     66/66

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:80
posted:3/30/2008
language:English
pages:66