Novel information management approaches for the development of

Novel information management approaches for the development of knowledge bases to support systems biology Prof. dr. Antoine van Kampen Bioinformatics Laboratory Academic Medical Center (AMC) NCSB meeting February 26, 2009 Bio. Systems Data Analysis Group University of Amsterdam

Data, Information, Knowledge and Wisdom Connectedness Wisdom Knowledge Information Data Understanding principles Understanding patterns Understanding relations Understanding

Data, Information, Knowledge and Wisdom Knowledge Information Data

Data, Information, Knowledge and Systems Biology Knowledge Information Data

Bringing data, information and knowledge towards application Bio. Expert implements knowledge bases

Experts already represent their knowledge… Human readable, not computer processable Annu. Rev. Biochem. 2006. 75: 295– 332

Fatty acid metabolism from KEGG: many facts but little insight

Different substrates use different enzymes Peroxisome Thiolysis Background knowledge: what is Hydration? dehydrogenation hydration mitochondrion Oxidation Peroxisomal beta-oxidation ceases at Octanoyl Co. A Activation

Bio. Expert project overview Bio. Expert Information Management Framework Knowledge Base KB may contain any type of biomedical knowledge; not restricted to pathways

Bio. Expert knowledge bases Bio. Expert Framework Peroxisome (Px. KB) Polyphenol degradation (Pp. DKB) NCSB-NBIC (open position) Yeast Glycolysis (Yg. KB) Metabolic Syndrome (Me. SKB) Systems Biology Start a. s. a. p Yg. KB: Prof. dr. B. Teusink (Molecular Cell Physiology, VU) Me. SKB: Dr. K. Willems van Dijk (Human and Clinical Genetics, LUMC) Jeroen Jeneson (Biomedical NMR, TUe), Roeland Merks (CWI), Hans van Beek (VU)

The Peroxisome Knowledge base (Px. KB)

What are peroxisomes? peroxisome Function Fatty acid β-oxidation Bile acid formation Fatty acid α-oxidation Plasmalogen biosynthesis Cholesterol biosynthesis? Glyoxylate breakdown other Disorders Peroxisome biogenesis disorders Zellweger syndrome Single peroxisomal enzyme deficiencies X-linked adrenoleukodystrophy Acyl-Co. A oxidase deficiency D-bifunctional protein deficiency 2 -methylacyl-Co. A racemase deficiency Rhizomelic chondrodysplasia punctata type 2 Rhizomelic chondrodysplasia punctata type 3 Refsum's disease Hyperoxaluria type 1 Glutaric acidemia type 3 (Mevalonate kinase deficiency? ) Acatalasaemia Mulibrey nanism ALDP ACOX 1 D-BP AMACR DHAPAT ADHAPS Phy. H AGT ? ? ? MVK CAT MUL

The challenges • Knowledge representation • Knowledge acquisition • Knowledge navigation and querying • Knowledge application

Knowledge representation • Semantic Web Initiative: – represent web content in a form that is more easily machineprocessable – to use techniques to take advantage of these representations • Semantic Web Aim: – more advanced knowledge management systems Antoniou, Van Harmelen. A Semantic Web Primer, 2 nd edition. Cambride: The MIT Press, 2008.

Semantic Web technology RDF Resource Description Framework Provides foundation for representing and processing metadata RDF Schema Primitive ontology language SKOS Simple Knowledge Organization System An ontology for expressing the basic structure and content of e. g. , controlled vocabularies. OWL Web Ontology Language Proposed standard for web-ontologies. Allows to describe the semantics of knowledge in a machine-accessible way SPARQL RDF query language

Foundation: RDF triple Graphic representation of triple: Subject Mitochondrion property part-of Object Cell

The challenges • Knowledge representation • Knowledge acquisition • Knowledge navigation and querying • Knowledge application

Knowledge presentation: Concept Maps • Intuitive – Non-technical – “Easy” to make • Clear – Graphical presentation • Concise – Focus / context A. J. Canas et al. , “Concept maps: Integrating knowledge and information visualization, ” in , vol. 3426, Lecture Notes in Computer Science (BERLIN: Springer-Verlag Berlin, 2005), 205 -219.

Did you recognize the RDF triples?

Semantic typing (RDFS / OWL)

The challenges • Knowledge representation • Knowledge acquisition • Knowledge navigation and querying • Knowledge application

The challenges • Knowledge representation • Knowledge acquisition • Knowledge navigation and querying • Knowledge application

Application of KB in systems biology • Annotation of biological system – Detailed and curated – Linked to external data(bases) – Possibly embedded in larger context (not part of modeling) – Facilitates mathematicians to correctly define model • Annotation of mathematical models – Allows biologists to verify model • Validation of results – Are results consistent with available knowledge • Generation of new hypotheses

(How to) connect Bio. Expert to. . . • SBML • Cell. ML • MIRIAM: Minimal Information Requested in the Annotation of Biochemical Models • Systems Biology Ontology • TEDDY: Terminology for the Description of Dynamics • MIASE: Minimum Information About a Simulation Experiment • Ki. SAO: Kinetic Algorithm Ontology • Sys. Mo DB / Taverna workflows / etc

The Bio. Expert Team Dr. Andrew Gibson Project leader Wiebe Posthuma Dr. Gerbert Jansen Ph. D position (translational medicine) Serge Barth, MSc Ph. D position Grey statistical models Joris Scharp Paul van Hooft Position NCSB-NBIC systems biology