Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Systems ReconstructionTM Technology CONFIDENTIAL Analysis of Proteomics Data in the Context of Human Systems Biology Tatiana Nikolskaya, PhD SCO & President, GeneGo, Inc April 22, 2004 Workshop II: Medical Applications and Protein Networks Copyright GeneGo 2000-2003 Why Systems Reconstruction? • CONFIDENTIAL • Avalanche of genomic, gene expression, proteomic, metabolomic data on one hand Complex human diseases and patient’s data on the other Understanding of the actual cause of complex diseases on molecular level is still at its infancy To the HT data mining boils down to statistical analysis Integration, visualization and data mining of each and every type of HT remains a BIG problem How to bring different types of molecular and clinical data together? • • • • • • • Systems Reconstruction Technology Pathway-centered database Clean and comprehensive content: human cellular pathways Necessary tools to manipulate each and every type of molecular data Identified links between pathway elements and human diseases Ability to retrieve disease-specific pathways Ability to identify and propose missing/unknown pathways from HT data • • • • Copyright GeneGo 2000-2003 CONFIDENTIAL MetaCore™ Platform SNP Analysis Tools Importing Tools Gene Expression Tools: Affy, Agilent, Resolver parsers Pathway Editor Network Tools SAGE analysis tools Visualization Tools Proteomics Tools Metabolomics Tools Proprietary Content of 23,000 curated pathway building blocks Novel Database Architecture Oracle Based 150 relational tables Copyright GeneGo 2000-2003 What makes MC special? • CONFIDENTIAL The most comprehensive DB of human pathways: – 23,000 human 1-step pathway blocks, rat and mouse orthologs – Vertical integration of pathways: from receptors to core effectors • Extensive manual pathway curation: – – – – • Associations with human diseases/conditions Tissue specificity, sub-cellular localization, effects, mechanisms >270,000 synonyms resolved Custom pathway editor to add/change the pathways Unique database architecture: – Pathways is the backbone for experimental data, literature info – Concurrent mapping of different types of HT data on maps and networks – Multiple time points, treatments/dosages, custom colors, custom ranges • Flexible visualization tools and options: – – – – – Mapping of customer’s data on maps and computer-generated networks Disease- and tissue-specific filters Tools for expanding and collapsing pathways on networks User’s choice between data sets and algorithms for generating networks Tools for reducing overall network’s complexity and expanding it Copyright GeneGo 2000-2003 MetaCoreTM: Content and Capabilities MetaCoreTM, a database of human metabolism and its regulation CONFIDENTIAL MetaCoreTM, data-mining tools, algorithms and visualization 23,000 pathway building blocks Metabolic and signaling networks >300 expert-curated maps Rat and Mouse orthologs 270,000 synonyms resolved Pathway editor 32,000 pathway/disease links 3,250 human diseases Transcriptional factors and sites 1,200 journals/37,000 unique ref Concurrent visualization of HT data Affymetrix and Agilent array parsers Proteomic and metabonomic parsers Custom and NCBI SAGE data parser Copyright GeneGo 2000-2003 CONFIDENTIAL MC: The Most Comprehensive Database on Human Biology Affymetrix chip ID U95Av2 HG-U133A HG-U133B GeneGo MetaCore ~82 % ~40 % GeneLogic GeneExpress ~20 % ~8 % SpotFire ~16% ~5% 75% of known human proteins can be visualized on maps and networks Copyright GeneGo 2000-2003 Pathways in MetaCore: Maps and Networks CONFIDENTIAL Interactive, static maps – >300 maps – Signaling, receptors, metabolism – Backbone of formalized “state of art” in the field Networks of protein interactions – Dynamic; built “on-the-fly” – Exploratory tool – Build new pathways for genes of interest Copyright GeneGo 2000-2003 CONFIDENTIAL Why Manual Curation? Microarray A Source of interaction NLP associations MC curated networks Two hybrid screen 2D gel B C D P interaction Confidence in a 4 step pathway .25 .4% .5 6% .75 30% .99 96 % Copyright GeneGo 2000-2003 Projected number of 5-step pathways 1 2 3 4 CONFIDENTIAL 5 More than 2,000,000,000 5-step pathways Copyright GeneGo 2000-2003 Maps and Networks Legend CONFIDENTIAL Copyright GeneGo 2000-2003 MetaCoreTM: Depth and precision… CONFIDENTIAL 1. transcription 2. processing RNA 3. transport PNA from nucleus 4. stabilization of RNA 5. translation 6. Protein transport 7. Folding & stabilization 8. allosteric modification 9. covalent modification Copyright GeneGo 2000-2003 MC: Reconstruction of “Vertical” Pathways from Receptors to Core Effectors CONFIDENTIAL adenosine Copyright GeneGo 2000-2003 Unique Capabilities: Visualization of Different Types of HT Data within the Same System Gene expression CONFIDENTIAL Protein levels Protein Interactions Metabolite concentrations Data parsers link user’s data to relevant molecular objects in MetaCore DB Copyright GeneGo 2000-2003 Concurrent Visualization of Different HT data Agilent Affymetrix Proteomic CONFIDENTIAL SAGE Copyright GeneGo 2000-2003 CONFIDENTIAL MC : Relevance to Human Diseases Breast cancer Alzheimer disease lung carcinoma Alzheimer disease chronic Leukemia breast cancer Cancer Breast cancer Brain tumors adenocarcinoma Tangier disease breast cancer Alzheimer disease neuroblastoma T cell lymphoma Breast cancer bladder transitional cell carcinoma bladder cancer lung cancer squamous cell carcinoma Alzheimer disease diabetes mellitus Alzheimer disease colorectal carcinoma breast cancer Parkinson disease pancreatic cancer glioblastoma Alzheimer disease chronic myelogenous leukemia Wiskott-Aldrich syndrome Friedreich ataxia Alzheimer disease Parkinson disease Prostate cancer nephronophthisis 1 Alzheimer disease Alzheimer disease Parkinson disease B-cell chronic lymphocytic leukemia Friedreich ataxia BorjesonForsmanLehmann syndromecondent ial cerebellar hypoplasia Acute myelogenous leukemia Creutzfeldt-Jakob disease Parkinson disease bladder cancer Lupus erythematosus polycystic kidney disease cystic fibrosis Alzheimer disease squamous cell carcinoma hepatocellular carcinoma Bladder cancer Copyright GeneGo 2000-2003 MC Value: HT Data Mining. Maps Inhibitors upregulated in Down-regulated Glaucoma in Glaucoma CONFIDENTIAL Copyright GeneGo 2000-2003 MC value: HT Data Mining. Networks CONFIDENTIAL Copyright GeneGo 2000-2003 CONFIDENTIAL MC value: Identify Novel Pathways for Your Genes Copyright GeneGo 2000-2003 User Chooses Network-creating Algorithms CONFIDENTIAL Copyright GeneGo 2000-2003 Generate Networks From the List of Genes/Proteins. CONFIDENTIAL Through Internal transcriptional database Copyright GeneGo 2000-2003 Generate Networks around Single Gene/Protein CONFIDENTIAL Copyright GeneGo 2000-2003 Custom Ranges and Colors CONFIDENTIAL 1. Select experiment 2. Show all descriptions 3. Add ranges 4. Switch to indicator state 5. Select ranges to visualize Copyright GeneGo 2000-2003 Pathway Editor: view/edit Interactions CONFIDENTIAL 1. Search for a protein class 2. View its interactions 3. Edit interactions • Edit effect • Edit mechanism • Insert references Copyright GeneGo 2000-2003 Pathway editor: adding new interactions CONFIDENTIAL 1. Search for new interaction’s vertices and their existing interactions 2. Edit existing existing interactions (if any) or add a new one 3. Use pull-down menus to specify effect and mechanism (if known) 4. Add references Copyright GeneGo 2000-2003 Hardware requirement, Maintenance and Upgrades • CONFIDENTIAL Server – Oracle 8 DBMS or higher, Apache web server with mod_perl – 2 or more P4/XEON CPU’s with 512-1024 MB of RAM recommended – SCSI HDD is recommended – Linux OS 7.3 (RedHat) – MetaCore™ takes about 300MB of space • Client: – Internet Explorer 6.0 or higher is required – Macromedia Flash 6 (MX) Plug-In is required – P4 CPU and 256MB of RAM is recommended • Maintenance is included in the annual fee • Updates will be shipped quarterly • Web based, easy to use and access Copyright GeneGo 2000-2003