1. Sanofi big data approach for
translational medicine
MongoDB World 2014
New York
June 23-25, 2014
1

2. Agenda
● Sanofi a Pharma Company
● Translational Medicine Concept
● Sanofi Architecture View
● Why MongoDB?
● Use Cases: Overview and Deep Dive
● Benefits, Lessons Learned & Next Steps
2

3. Information of December 31st 2013
SANOFI GROUP
●We are a global
healthcare company
engaged in the
research, development,
manufacturing and
marketing of healthcare
solutions. present
in more than
100
countries
more than
110 000
employees
A
comprehensive
offer
of pharmaceuticals,
vaccines and innovative
therapeutic solutions
112
Industrial sites
in 41 countries
R&D
A major biopharmacy
player
• 45% of revenues
generated by biologics
• 80% of development
projects are biologics
€33 bn*
In sales in 2013
* €32,951 M
Sanofi: A Big Pharma Company

4. Sanofi R&D
(1) Investor Relations Annual Results Meeting dated April 29th, 2014, excluding Merial R&D
(2) Source : Document de Référence 2013.
(3) Source : Document de Référence 2013. Includes all R&D positions within the Group (Affiliates, Industrial Affairs, etc)
Billions invested in R&D in
2013(2)
4,7 €
worldwide(2)
20 sites
More than
Clinical Study Units in
40 countries
Molecules and vaccines
in the R&D portfolio
50
Including 12 in late stage(1)
Employees worldwide contributing to
the research and development of
innovative health solutions(3)
16 500
Approximately
4

5. Introduce ourselves
Global Solution Leader at Sanofi-Aventis R&D
• Translational Medicine Platform.
• Sanofi Representative in the eTRIKS
Project Manager at Sanofi-Aventis R&D
• Early research Enterprise Data Warehouse
• Big data project for Translational Medicine
Platform
Scrum Master, Technical leader at Apside
• Deep pharma experience (Servier, Fabre, Sanofi)
• Advanced agile expert
5

6. Pharma Company Mutation
From curative to preventive medicine
● Main challenges for Pharmas
● Competition by generic,
● End of the blockbuster age,
● Leak of innovation,
● New paradigm: 4 P’s concept & translational medicine
● Personalized,
● Predictive,
● Preventive,
● Participatory
6

7. Research
Clinical
Knowledge
OMICS
data
Hypotheses
Trials
EHRs
CROs
Evidence
● Efficient clinical trials
● Personalized medicine
● New capabilities for
OMICS data
● One stop shop for TM
data & knowledge
Translational Medicine Concept
Bridge Clinical & Research
7

8. New challenges raised by TM
● A diversity of
objects to be
connected
● Maintain
consistency &
traceability
● Extract knowledge,
gain understanding
8

9. Research
Molecular
Profiling
Clinical
Analysis
Big
Data
TM
Data-Hubs
Biomarker-
Research
Target validation
Interpretation
Biomarker
Epidemiology
Extraction
Curation
Normalization
Loading
Coordination & Collaboration
TM Data Mgmt Layer
A
C
B
D
011001101 BIOMARKER
MongoDB as a
central repository to
enhance biomarker
research
Sanofi Architecture View
Schematic Overview
9

10. Why MongoDB?
ScalableAdapted to file &
metadata
Easy
To use
To install
To understand
and to adopt 
10

11. How we moved to mongoDB?
White
paper on
Big data
Install,
benchmark
PoC
1st
Release in
PROD
Q42013Q12013
Wow
11

12. GridFS
fs.file/fs.chunk
Logs, tools…
User profile
Config
Metadata
Service
layer
(REST
API)
Web site
Third party software
Fuse/Desktop software
MongoDB Use Cases
General architecture
12

13. MongoDB Use Cases
Annotation
13

14. MongoDB Use Cases
Annotation
14

15. ● Geographic zone / Health activity
MongoDB Use Cases
360° Data Explorer
AND
7 docs
AND
16 docs
Same Data, exposed in
different organization
 Facetted Navigation
● Disease or Syndrome / Receptor
15

16. Lessons Learned & Challenges
● Link together the hybrid landscape
● Keep in sync MongoDB an the Solr index
● IT/infrastructure department needs supports
● MongoDB is not yet a standard in Sanofi
● But Convincing managers/architects is easy
• PoC can be set up very easily and show immediate
benefits
• The fast growing community/clients help a lot
16

17. Next Steps
● Curation process
● Improve curation by integrating open refine
● Collaboration
● Set up security at the smallest data piece (‘cell level
security’)
● Scale up
● Deploy to more department and geographic area
● Transition mongoDB support to infrastructure
teams
● To be finalized
17

18. Benefits
● Scientists:
● Time gain in tagging and curation
● Awareness of existing data
● Explore data
● Integration of external data
● IT:
● Time of development (agility) & implementation
● Flexibility
● Performance
● Documentation, support, training, MOOC
18

19. Question and answers
19

20. Draft Agenda
● Challenges & business case (2 slides), the sanofi TM4P platform (2 slides):
how sanofi can be more innovative
● General idea (2 slides): research data hub : transform the way it’s done today
● Implementation
● MongoDB as a central store (metadata, gridFS: 2 – 3 slides)
● Ecosystem: solR integration (1 slide), DML (1 slide)
● (Business track not too technical) system data flow
● Demo by screenshots ? (3 slides)
● Challenges: set up an hybrid solution (solR/xxx )
● User testimony benefits (time, data quality & understanding)
● IT testimony: reduce the development time/agile
● deployment plan
● Next step/perspectives (2 slides)
20

21. New challenges raised by TM
21

22. 22

23. 23

24. 24

25. DML
• STORE
• ORGANISE
• COLLABORATE
• LINK DATA
• VISUALIZE
• ANALYSE
25