BioUML integrated platform for building virtual cell and virtual physiological human Fedor Kolpakov Institute of Systems Biology Laboratory of Bioinformatics, Design Technological Institute of Digital Techniques Novosibirsk, Russia

BioUML platform BioUML is an open source integrated platform for systems biology that spans the comprehensive range of capabilities including access to databases with experimental data, tools for formalized description, visual modeling and analyses of complex biological systems. Due to scripts (R, JavaScript) and workflow support it provides powerful possibilities for analyses of high-throughput data. Plug-in based architecture (Eclipse run time from IBM is used) allows to add new functionality using plug-ins. BioUML platform consists from 3 parts: BioUML server – provides access to biological databases; BioUML workbench – standalone application. BioUML web edition – web interface based on AJAX technology;

BioUML main features Supports access to main biological databases: –catalolgs: Ensembl, UniProt, ChEBI, GO… –pathways: KEGG, Reactome, EHMN, BioModels, SABIO-RK, TRANSPATH, EndoNet, BMOND… Supports main standards used in systems biology: SBML, SBGN, CellML, BioPAX, OBO, PSI-MI… database search: –full text search using Lucene engine –graph search graph layout engine visual modeling: –simulation engine supports (ODE, DAE, hybrid,1D PDE); –composite models; –agent based modeling; –parameters fitting; genome browser (supports DAS protocol, tracks import/export); data analyses and workflows – specialized plug-ins for microarray analysis, integration with R/Bioconductor, JavaScript support, interactive script console.

BioUML workbench

BioUML web edition

BioUML web edition dedicated Amazon EC2 server:

Text search: universal full text search engine based on Apache Lucene technology

Metaphor: biological systems reconstruction as solitaire (patience) game Desk – BioUML editor Solitaire – biological pathway Cards – biological objects (genes, proteins, lipids, etc.) Pack of cards – different biological databases

Graph layout

Visual modeling

Pane: simulation

Reports (templates)

Parameters fitting

Parameters fitting: main features Experimental data – time courses or steady states expressed as exact or relative values of substance concentrations Different optimization methods for analysis Multi-experiments fitting Constraint optimization Local/global parameters Parameters optimization using java script

Parameters fitting – user interface

Comparison with COPASI (10,000 simulations) MethodBioUML (4 cores) BioUML (1 core) COPASI (1 core) Evolutionary Programming –– 1 min 58,2sec 1 min 31,3 sec 1 min 16,6 sec Particle swarm7,1 sec 7,7 sec 6,9 sec 22,4 sec 15,3 sec 22,5 sec 1 min 32 sec 1 min 26,4 sec 1 min 07,1 sec Stochastic Ranking Evolution Strategy 7,5 sec 7,47 sec 6,9 sec 23,4 sec 23,5 sec 22,2 sec 1 min 25,0 sec 1 min 5,6 sec 1 min 8,8 sec Cellular genetic algorithm 7,7 sec 7,5 sec 7,2 sec 25,5 sec 22,1 sec 20,8 sec –

Genome browser

uses AJAX and HTML5 technologies interactive - dragging, semantic zoom tracks support Ensembl DAS-servers user-loaded BED/GFF/Wiggle files Genome browser: main features

Data analyses: collaboration & reproducible research

R world Java/BioUML world JavaScript host objects allows to merge R/Bioconductor and Java/BioUML worlds

Analysis workflow

Composite models Poster: Modular Approach To Modeling Of The Apoptosis Machinery Agent based modeling Poster: Agent based modelling – plug-in for BioUML platform

Composite model of apoptosis (286 species, 684 reactions)

Bentele M, 2004 Neumann L, 2010 CD95L module and results of fitting its dynamics to experimental data

Agent based model of arterial hypertension (blood pressure regulation)

October 10: Tutorial 3: BioUML – integrated platform for building virtual cell and virtual physiological human October 12, , Parallel session: Fedor Kolpakov: Modular Approach To Modeling Of The Apoptosis Machinery October 11-12, poster session BioUML – Integrated Platform for Building Virtual Cell and Virtual Physiological Human Modular Approach To Modeling Of The Apoptosis Machinery The Optimization Plug-in For The BioUML Platform Agent based modelling - plug-in for BioUML platform Numerical Analysis Of The Complex Model Of Human Cardio- Vascular System Using 1D Hemodynamic Model Stand of geneXplain GmbH company

Acknowledgements Part of this work was partially supported by the grant: European Committee grant № “Net2Drug” European Committee grant № “LipidomicNet” Integration and interdisciplinary grants №16, 91 of SB RAS. BioUML team Software developers Biologists Nikita Tolstyh Ilya Kiselev Ruslan Sharipov Tagir Valeev Elena KutumovaIvan Yevshin Anna Ryabova Alexey Shadrin