This document describes BiVeS and BudHat, tools for version control of computational models. BiVeS detects differences between versions of models encoded in standardized formats like SBML. It maps common elements and constructs an XML diff file. BudHat visualizes these diffs, highlighting changes in reaction networks or providing a readable report. The tools are open source and intended to extend existing model repositories with version control capabilities. A demonstration of BiVeS and BudHat in action is provided.

1. SYSTEMS BIOLOGY
BIOINFORMATICS
ROSTOCK
S E Ssimulation experiment management system
BiVeS & BudHat
Version Control for Computational Models
MARTIN SCHARM
Department of Systems Biology & Bioinformatics
Faculty of Computer Science & Electrical Engineering
University of Rostock
http://sems.uni-rostock.de
HARMONY 2013
Farmington, CT, USA
May 20, 2013 Bives & Budhat | Martin Scharm 1 / 13

2. SYSTEMS BIOLOGY
BIOINFORMATICS
ROSTOCK
S E Ssimulation experiment management system
track development
store retrieve
rank
Management
Δ
Δ
Version 1
Version 2
latest
Format-independent,
graph-based storage
Information Retrieval-based
search and ranking
Diff-based
version control
http://sems.uni-rostock.de/
May 20, 2013 Bives & Budhat | Martin Scharm 2 / 13

3. Version Control
Intro
• many models
• multiple versions of each model
• several repositories
time
#models
May 20, 2013 Bives & Budhat | Martin Scharm 3 / 13

4. Version Control
Intro
• many models
• multiple versions of each model
• several repositories
version x-1 version x version x+1
May 20, 2013 Bives & Budhat | Martin Scharm 3 / 13

5. Version Control
Intro
• many models
• multiple versions of each model
• several repositories
May 20, 2013 Bives & Budhat | Martin Scharm 3 / 13

6. Version Control
How to get versions
• the FTP way
• the VCS way
usr@srv % ftp ftp.ebi.ac.uk
Connected to ftp.gslb.ebi.ac.uk.
220-
[...]
230 Login successful.
Remote system type is UNIX.
Using binary mode to transfer files.
ftp> cd pub/databases/biomodels/releases/2012-08-11/
250 Directory successfully changed.
ftp> get BioModels_Database-r24_pub-sbml_files.tar.bz2
local: BioModels_Database-r24_pub-sbml_files.tar.bz2 remote:
BioModels_Database-r24_pub-sbml_files.tar.bz2
200 PORT command successful. Consider using PASV.
150 Opening BINARY mode data connection for BioModels_Database-r24_pub-
sbml_files.tar.bz2 (91772 bytes).
226 Transfer complete.
May 20, 2013 Bives & Budhat | Martin Scharm 4 / 13

7. Version Control
How to get versions
• the FTP way
• the VCS way
usr@srv % hg clone http://models.cellml.org/w/andre/nickerson-2008
destination directory: nickerson-2008
requesting all changes
adding changesets
adding manifests
adding file changes
added 2 changesets with 131 changes to 131 files
updating to branch default
131 files updated, 0 files merged, 0 files removed, 0 files unresolved
usr@srv % hg log path/to/model
[...]
May 20, 2013 Bives & Budhat | Martin Scharm 4 / 13

8. Version Control
How to get versions
• the FTP way
• the VCS way
version x-1 version x version x+1
track versions
what happened ?
May 20, 2013 Bives & Budhat | Martin Scharm 4 / 13

9. Version Control
good news
A r C
B
D
cycE/cdk2
RB/E2F
RB-Hypo
free E2F
A r
B
C
D
E s
RB/E2F
RB-Hypo
free E2F
cycE/cdk2
RB-Phos
new insights
Waltemath et al.: Improving the reuse of computational models through version
control. Bioinformatics (2013) 29(6): 742-728;
May 20, 2013 Bives & Budhat | Martin Scharm 5 / 13

10. BiVeS
Difference Detection
A r C
B
D
cycE/cdk2
RB/E2F
RB-Hypo
free E2F
A r
B
C
D
E s
RB/E2F
RB-Hypo
free E2F
cycE/cdk2
RB-Phos
A
r
B
C
D
A
r
B
C
D
E
s
Biochemical Model Version Control System
• compares models encoded in standadized
formats (currently: and )
• maps hierarchically structured content
mapping
May 20, 2013 Bives & Budhat | Martin Scharm 6 / 13

11. BiVeS
Mapping
• input: 2 XML documents
• map unambiguos
nodes/subtrees
• propagate mapping up/down
• exploit further knowledge of
biol. systems
same ID
same
subtree
hash
...
... ...... ...
...
... ...... ...
May 20, 2013 Bives & Budhat | Martin Scharm 7 / 13

12. BiVeS
Mapping
• input: 2 XML documents
• map unambiguos
nodes/subtrees
• propagate mapping up/down
• exploit further knowledge of
biol. systems
eq. netw....
... ...... ...
...
... ...... ...
May 20, 2013 Bives & Budhat | Martin Scharm 7 / 13

13. BiVeS
Mapping
• input: 2 XML documents
• map unambiguos
nodes/subtrees
• propagate mapping up/down
• exploit further knowledge of
biol. systems
A B
C
D E
A B
C
D E
A B
C
D E
May 20, 2013 Bives & Budhat | Martin Scharm 7 / 13

14. BiVeS
Difference Detection
A r C
B
D
cycE/cdk2
RB/E2F
RB-Hypo
free E2F
A r
B
C
D
E s
RB/E2F
RB-Hypo
free E2F
cycE/cdk2
RB-Phos
A
r
B
C
D
A
r
B
C
D
E
s
Biochemical Model Version Control System
• compares models encoded in standadized
formats (currently: and )
• maps hierarchically structured content
• constructs a diff (in XML format)
<XML>
Diff
moves
product of r: C
deletes
product of r: B
inserts
species: E
product of r: E
reaction s
</XML>
mapping
di construction
May 20, 2013 Bives Budhat | Martin Scharm 8 / 13

15. BudHat
Diff Visualization
A r C
B
D
cycE/cdk2
RB/E2F
RB-Hypo
free E2F
A r
B
C
D
E s
RB/E2F
RB-Hypo
free E2F
cycE/cdk2
RB-Phos
A
r
B
C
D
A
r
B
C
D
E
s
XML
Diff
moves
product of r: C
deletes
product of r: B
inserts
species: E
product of r: E
reaction s
/XML
• calls BiVeS to construct the diff
• displays the result in various formats
• the XML diff
• a reaction network highlighting the
changes using
• a human readable report
A r B
C
D
E s
May 20, 2013 Bives Budhat | Martin Scharm 9 / 13

16. BiVeS BudHat
DEMO
lets take a look at our tools in action!
http://budhat.sems.uni-rostock.de
May 20, 2013 Bives Budhat | Martin Scharm 10 / 13

17. BiVeS BudHat
Integration
BiVeS BudHat can be integrated in extisting repositories!
Requirements:
• Models must be encoded in standadized formats (e.g. CellML or SBML)
• Each version of a model must be accessible for BudHat (either physically or
through an API)
• Java based web server (e.g. tomcat), may run on a seperate node
If you are interested in using our tools feel free to contact us.
May 20, 2013 Bives Budhat | Martin Scharm 11 / 13

18. BiVeS BudHat
Summary
• BiVeS = Difference detection for hierarchical structures
• BudHat = Visualization of changes
• Both tools are open source
• We do not want to establish yet another platform
• Our goal is to extend existing models repositories with a valuable version
control mechanism
May 20, 2013 Bives Budhat | Martin Scharm 12 / 13

19. SYSTEMS BIOLOGY
BIOINFORMATICS
ROSTOCK
S E Ssimulation experiment management system
That’s it! Stay tuned ;-)
@SemsProject
http://sems.uni-rostock.de
http://budhat.sems.uni-rostock.de
Questions? Suggestions? Recommendations? Drop me an email:
martin.scharm@uni-rostock.de
May 20, 2013 Bives Budhat | Martin Scharm 13 / 13