Lecture at the LCSB 2017
•
0 likes•9 views
Building the biological Babylon Tower with standard representations of pathways
Recommended
Recommended
More Related Content
Similar to Lecture at the LCSB 2017
Similar to Lecture at the LCSB 2017 (20)
More from Nicolas Le Novère
More from Nicolas Le Novère (10)
Recently uploaded
Recently uploaded (20)
Lecture at the LCSB 2017
- 1. Building the biological Babylon Tower with Building the biological Babylon Tower with standard representations of pathways standard representations of pathways Nicolas Le Novère Babraham Institute, n.lenovere@gmail.com
- 2. Biochemical pathways are old … Nicholson (1970) Michal (1984) Gortner (1949) “Hand drawing” on paper → no software-based browsing, processing and analysis
- 3. Biochemical pathways are old … or not so much ● 1990s: high-throughput data generation → Large amount of knowledge increase in computing power → automatic reconstruction, browsing and analysis ● Databases: EcoCyc (1994), KEGG (1995), Reactome (2000) ● Formats: BioPAX (2000), SBML (2000), PSI-MI (2002) Nicholson (1970) Michal (1984) Gortner (1949)
- 5. Oda K. Mol Syst Biol. 2005;1:2005.0010. UNDERSTAND Signalling pathways
- 6. Davidson's map Davidson et al. Science (2002) R.I.P. September 2015 UNDERSTAND Gene regulatory networks
- 7. Liu et al (2009) Nat Rev Drug Discov CURE
- 8. Liu et al (2009) Nat Rev Drug Discov Breast cancer CURE
- 9. Abou-Sleiman et al (2006) Nat Rev Neurosci Parkinson’s disease CURE
- 10. Abou-Sleiman et al (2006) Nat Rev Neurosci Parkinson’s disease CURE
- 12. Team Nanjing, iGEM (2013) Synthetic Biology BUILD
- 13. Ambiguity of usual representations
- 14. Ambiguity of usual representations
- 15. Can-this be understood by biologists?
- 16. Can-this be understood by biologists? Stimulates? but ... what exactly? Associates into? Translocates? No idea. Reciprocal stimulation? Is degraded? Stimulates gene transcription?
- 17. Ambiguity of usual representations
- 18. Ambiguity of usual representations
- 20. Every computer scientist understands those
- 21. Every physicist understands this
- 22. Every engineer understands that
- 23. The Systems Biology Graphical Notation A tool to graphically describe and unambiguously interpret biochemical and cellular events Limited amount of symbols Smooth learning curve Can represent logical or mechanistic models, biochemical pathways, at different levels of granularity Detailed technical specification, precise data-models, standard API and growing software support Developed over ten years by a diverse community, including biologists, modellers, computer scientists etc.
- 25. Different granularities of a pathway ● No temporal sequence ● No directionality ● No biochemical effects ● No mechanistic descriptions ● No temporal sequence ● No directionality ● No biochemical effects ● No mechanistic descriptions
- 26. ● Directionality of influence ● Directionality of influence
- 27. ● Directionality of influence ● Directionality of effect ● Directionality of influence ● Directionality of effect
- 28. ● Directionality of influence ● Directionality of effect ● Biochemical effect ● Directionality of influence ● Directionality of effect ● Biochemical effect
- 29. One notation – three languages
- 30. Logical modelling Signalling pathways, gene regulatory networks Functional descriptions: the Activity Flows
- 34. Davidson's map Davidson et al. Science (2002)
- 35. SBGN Activity Flows L1 reference card
- 37. TIMTOWTDI (There Is More Than One Way To Do It)
- 41. *YOU* choose how to represent it SBGN makes sure it will be unambiguously interpreted *YOU* choose how to represent it SBGN makes sure it will be unambiguously interpreted
- 42. Puting in mechanisms: describing processes
- 43. A biochemical reaction is a process S1 S2 E P
- 44. A biochemical reaction is a process → Reconstruction of state variable evolution from process descriptions: ● Processes can be combined in ODEs (for deterministic simulations); transformed in propensities (for stochastic simulations) ● Systems can be reconfigured quickly by adding or removing a process S1 S2 E P
- 45. Process modelling Biochemistry, Metabolic networks Generally within “closed world” Subjected to combinatorial explosion SBGN Process Descriptions
- 46. Open world Closed world Anything not explicitly stated is unknown Failure to observe does not imply non-existence New pieces of knowledge do not affect prior pieces Anything not explicitly stated does not exist Failure to observe implies non-existence New pieces of knowledge might change the meaning of prior pieces
- 49. SBGN Process Diagram L1 reference card
- 51. Process Nodes
- 52. Connecting arcs
- 53. Metabolism in SBGN PD
- 54. Signalling in SBDG PD
- 57. All those diagrams are identical C C D D A A B B C D A B A B C D C D A B
- 58. Reactome
- 60. Recon map
- 61. PathVisio
- 62. CellDesigner
- 63. Vanted
- 64. Rule-based modelling Molecular Biology “Open world” Independent rules: no explosion Entity Relationships
- 67. SBGN Entity Relationships L1 reference card pre:label
- 68. SBGN Entity Relationships L1 reference card continuants, things that exists (or not) pre:label
- 69. SBGN Entity Relationships L1 reference card occurrents, events that may happen (or not) pre:label
- 70. A B T P Entity Relationships can be viewed as rules
- 71. A B If A exists, the assignment of the value P to the state variable T of B is increased Entity Relationships can be viewed as rules P T
- 72. A B If A exists, the assignment of the value P to the state variable T of B is increased Entity Relationships can be viewed as rules (A stimulates the phosphorylation of B on the threonine) P T
- 73. A B If A exists, the assignment of the value P to the state variable T of B is increased C S P Entity Relationships can be viewed as rules If P is assigned to the state variable T of B, the assignment of the value P to the state variable S of B is decreased T P
- 74. Multistate and combinatorial explosion EGFR Y654 Y721 Y845 Y891 Y920 Y992 Y1045 Y1068 Y1086 Y1101 Y1148 Y1173 Process Descriptions: “once a state variable value, always a state variable value” 212 = 4096 states (i.e. EPN glyphs) for EGFR and 4096 complexes between EGFR and targets P target1 P target2 P target3 P target4 P target5 P target6 P target7 P target8 P target9 P target10 P target11 P target12
- 75. Regulation of synaptic plasticity by calcium
- 77. SBGN AF L1V1.0 04-SEP-2009 SBGN AF L1V1.2 27-JUL-2015 SBGN ER L1V1.0 03-SEP-2009 SBGN ER L1V1.1 06-OCT-2010 SBGN ER L1V1.2 14-APR-2011 SBGN ER L1V2 08-AUG-2015 SBGN PD L1V1.0 15-AUG-2008 SBGN PD L1V1.1 02-SEP-2009 SBGN PD L1V1.2 03-OCT-2010 SBGN PD L1V1.3 14-FEB-2011 Decision to Create SBGN OCT-2005 Tokyo SBGN 1 FEB-2006 Tokyo SBGN 2 OCT-2006 Yokohama SBGN 2.5 MAR-2007 Heidelberg SBGN 3 SEP-2007 Long Beach SBGN 3.5 JAN-2008 Okinawa SBGN 4 OCT-2008 Rostock SBGN 4.5 APR-2009 Waiheke SBGN 5 SEP-2009 Stanford SBGN 5.5 APR-2010 Wittenberg COMBINE OCT-2010 Edinburg HARMONY APR-2011 New-York COMBINE SEP-2011 Heidelberg HARMONY MAY-2012 Maastricht COMBINE AUG-2012 Toronto SBGN 9 MAY-2013 Edinburgh COMBINE AUG-2014 Los Angeles HARMONY MAY-2013 Farmington COMBINE SEP-2013 Paris HARMONY APR-2014 Manchester HARMONY APR-2015 Wittenberg COMBINE OCT-2015 Salt Lake City HARMONY JUN-2016 Auckland COMBINE SEP-2016 Newcastles
- 78. SBGN-ML
- 80. Are we done? Everything is solved, right?
- 82. Easy in AF
- 83. Cumbersome in ER
- 84. molecular species (entity pool nodes) belong to a single compartment impossible In PD
- 85. Multi-compartment complexes do not solve the problem, since they represent Non-covalent assemblies Loose the fact that Ephrin-A binds to Extracell part Loose the fact that Eph-A is a Multi-domain molecule
- 86. ● New glyph: “Domain”; generalisation of the current “Nucleic acid feature”. ● Domains can be associated in multi-domain entities pool nodes, the location being carried by the domains ● Allow complexes to be compartment-less ● Move content of complex from “decorations” to bona fide entity pool nodes
- 87. Hybrid maps
- 88. Hybrid maps
- 89. Use Submaps: PD in AF Submap in Process Descriptions Main map in Activity Flows
- 90. Submap in Activity Flows Main map in Process Descriptions Use Submaps: AF in PD
- 91. KT KR L Lc Allosteric regulations Fluxes Intensities and amounts
- 93. Creation, movement, destruction Is this a “pool” of identical entities?
- 94. Creation, movement, destruction Are those different “states” of the vesicle?
- 95. Creation, movement, destruction Are those different compartments?
- 96. Creation, movement, destruction Should those be “cloned”
- 97. Temporal order and Dynamics? Multi-scale? Topologies? Data integration?
- 99. SBGN editors Tobias Czauderna Emek Demir Ugur Dogrusoz Robin Haw Michael Hucka Hiroaki Kitano Nicolas Le Novère Augustin Luna Alexander Mazein Huaiyu Mi Stuart Moodie Falk Schreiber Anatoly Sorokin Alice Villéger Other main contributors Mirit Aladjem Frank Bergman Sarah Boyd Laurence Calzone Mélanie Courtot Tom Freeman Akira Funahashi Ralf Gauges Peter Ghazal Martijn van Iersel Hideya Kawaji Douglas Kell Sohyoung Kim Kurt Kohn Fedor Kolpakov Lu Li Yukiko Matsuoka Sven Sahle Chris Sanders Herbert Sauro Esther Schmidt Jacky Snoep Vasundra Toure Dagmar Waltemath Steve Watterson Katja Wegner Sarale Wimalaratne Guanming Wu THE COMMUNITY OF SYSTEMS BIOLOGISTS AND TOOL DEVELOPERS