tranSMART Community Meeting 5-7 Nov 13 - Session 2: Creating a Comprehensive Clinical and 'Omics Information Commons on Autism

tranSMART Community Meeting 5-7 Nov 13 - Session 2: Creating a Comprehensive Clinical and 'Omics Information Commons on Autism
Paul Avillach, Harvard University

1. Creating a comprehensive
clinical and 'omics information
commons on autism
Paul%Avillach,%MD,%PhD%
Harvard&Medical&School&
HEGP&Hospital,&Paris&–&Paris&Descartes&University&
INSERM&UMRS&872&eq&22&
Erasmus&MC&University&

2. Toward Precision Medicine: Building a Knowledge Network for
Biomedical Research and a New Taxonomy of Disease
Report from National academy of science, USA, 2011
2

3. HEGP background
Hôpital Laennec
(1634)
Hôpital Boucicaut Hôpital Broussais
Opening : July 2000-

4. HEGP BDW EHR/BDW integration

EHR :
Operational
Database
(ODS)

EHR :
Mirrored
Database
Production environment
Evaluation/Research
environment
ETL suite (Talend Open Studio)

Biomedical Data
Warehouse
(BDW)
External
Databases
Data Analysis
Data Mining
R
i2b2/tranSMART tools
Business Object
IBM Ilog Rules
Real time requests
Pr Patrice Degoulet CIO

5. PI: Isaac Kohane – Harvard - CBMI

6. HEGP i2b2

7. HEGP CDW use i2b2 CDW queries

8. +’omics
CDW%
Health care
Health Information System
Clinical Research
Pathology
HEGP
DRG
Biology
EHR forms
Imaging
Browser tools
available
for MD & Pharm
EHR reports i2b2
‘omics
ETL once a week
‘omics data
ETL
Structured data
from research
studies
ETL
Rx
BDW
Analysis tools %

9. • Integrated platform to support translational research
• Initiated by Johnson & Johnson et Recombinant 5 years ago
• Open-source since January, 2012
• Installed at HEGP since May, 2012
• Today, driven and maintained by the tranSMART
Foundation & communitie
hEp://transmarHoundaIon.org&
HEGP BDW

10. Objectives :
1. Integration of clinical, biological and ‘omics
data in one place – hypothesis free –
2. Generation of hypothesis by Clinicians /
Researchers

11. !
h0p://www.transmartproject.org%
“Omics” data integration

12. 12
L&Szalma&S,&Koka&V,&Khasanova&T,&Perakslis&ED.&EffecIve&knowledge&management&
in&translaIonal&medicine.&J&Transl&Med.&2010;&8:68.&

13. Pilot Study
HEGP - Paris

15. HEGP:&Canuel&V,&Avillach&P&

16. HEGP:&Canuel&V,&Avillach&P&

17. HEGP:&Canuel&V,&Avillach&P&

18. • R module in
tranSMART
• Published figure in
JCO
HEGP:&Canuel&V,&Avillach&P&

19. Phenotypic augmentation

20. Phenotypic augmentation

21. Type of Data
Electronic Health Record
Head and Neck Study
Head and Neck
Cancer Study
Treatment
Vital Status
Cinical Data
Biology
Patology
Death
Re-intervention
CancerrecurrenceHead and Neck Cancer
Studysprecific samples
tranSMART + i2b2 = Phenotypic augmentation
HEGP:&Joubert&F,&Avillach&P&

22. 06/05/2013
Badoual Cancer Res. 2013
p%=%0,04%
HEGP:&Joubert&F,&Avillach&P&

23. p%=%0,07%
p%=%0,03%
AVG&:&10&months&of&
addiIonal&follow&up&
with&i2b2&data&
HEGP:&Joubert&F,&Avillach&P&

24. PheWAS study
on i2b2/tranSMART
Neuraz A, Chouchana L, Malamut G, Le Beller C, Roche D, Beaune P,
Degoulet P, Burgun A, Loriot MA, Avillach P.

Phenome-wide association studies on a quantitative trait: Application to
TPMT enzyme activity and thiopurine therapy in pharmacogenomics.

PLoS Comput Biol. 2013 in press
24

27. Workflow

28. Phenotype%
Thiopurine%%
dose%
Low%
acHvity%
Intermediate%
acHvity%
Normal%
AcHvity%
10%%%
dose%
30%–%70%%%
dose%
100%%%
dose%
FDA & EMA
recommendations

?&
Methodes: Selection of trait:
enzymatic Activity TPMT

29. TPMT
HEGP:&Neuraz&A,…,&&Avillach&P&

30. ICD9LCM&
Mapping&
Very&High&TPMT&acIvity&VS&others&
HEGP:&Neuraz&A,…,&&Avillach&P&

31. HEGP:&Neuraz&A,…,&&Avillach&P&

32. Acknowledgments
HEGP
Informatics &
Public Health Dept :
• Eric Zapletal,
• Vincent Canuel,
• Antoine Neuraz,
• Fabien Joubert
Contact :
[email protected]
[email protected]
www.i2b2.org&
www.transmartproject.org&
Prof%Patrice%Degoulet,%past%CIO%
Prof%Anita%Burgun,%CIO%%

33. European Medical
Information Framework

34. • 58 partners
• 58 Million euros
• Started 1
st
jan 2013
• 5 year project
• PI: Bart Vannieuwenhuyse (Janssen) & Prof. Simon
Lovestone (KCL)
• PI Platform: Prof. Johan van der Lei
• Detect new biomakers:
– prediposition Alzheimer's disease
– Metabolic complication diabetes
EMIF project: European Medical Information Framework

35. 3 projects, 3 datasources:
• EHR derived patient data: 48 millions patient records :
observational studies (Jerboa tool)
• Cohort clinical + ‘omics data from : tranSMART option
– AD
– Metabolic :
~10 000 patients
• WP11 : Semantic harmonization
• Same concepts across all database : ETL process
Jerboa & tranSMART
EMIF project: European Medical Information Framework

36. Toward Precision Medicine: Building a Knowledge Network for
Biomedical Research and a New Taxonomy of Disease
Report from National academy of science, USA, 2011
36

38. ASDs SSC Cohort to BCH
Type% NaHonal%SSC%Lou%Kunkel%SSC%Link%to%BCH%i2b2%
AuIsm& 1&899& 356& 186&
Pdd& 549& 314& 163&
Nonspectrum& 0& 34& 20&
Aspergers& 275& 43& 17&
ReEs& 0& 1& 0&
TypicalLnoLdx& 0& 153& 93&
NA& 2& 19& 18&
Total% 2%725% 920% 497%
Probands)+)sibblings)

39. #%paHents%#%observaHons%
Vital:Weight& 367& 4&727&
Vital:Height& 359& 7&220&
SER& 421& 17&915&
Radiology&Procedures& 174& 1&142&
Procedures&(CPT)& 209& 1&470&
MedicaIons&(PrescripIon)& 212& 3&171&
Pathology&Repports& 220& 2&255&
Clinical&Notes& 419& 33&978&
MedicaIons& 189& 19&321&
Lab&tests& 360& 72&433&
Insurance&Payors& 375& 8&869&
MedicaIons&(History)& 217& 835&
Diagnosis&(ICD9)& 421& 26&323&
Clinic& 421& 14&169&
AMED& 164& 7&142&
Allergy& 246& 583&
AVG%=%298% %TOTAL%=%221%553%
ASDs SSC Cohort to i2b2 BCH

40. Autism information commons
From SSC National cohort to BCH i2b2

41. 41

42. 42

46. Dra]%