1 of 78

Facilitating integrative research

(EMBL-EBI)

Data!

2 of 78

Major resources and platforms

Multi-Omics

4 of 78

Synapse

Includes:

Accelerating Medicines Project – Alzheimer’s disease (AMP-AD)

CommonMind Consortium (CMC)

The Molecular Mechanisms of the Vascular Etiology of Alzheimer’s Disease (M2OVE-AD)

PsychENCODE

The Cancer Genome Atlas (TCGA)
Others...

https://www.radc.rush.edu/

5 of 78

Psychiatric Genomics Consotrium

6 of 78

NIMH Repository & Genomics

221 Studies - full range of illnesses affecting mental health

7 of 78

(Nature 2021)

8 of 78

Major resources and platforms

Result Browsers

9 of 78

The GWAS Catalog

17 of 78

PanUKBB Consortium

18 of 78

PanUKBB Consortium

19 of 78

Major resources and platforms

Neuroimaging

20 of 78

http://adni.loni.usc.edu/

Data from the North American ADNI’s study participants, including Alzheimer’s disease patients, mild cognitive impairment subjects and elderly controls.

ADNI

21 of 78

PNC

http://www.med.upenn.edu/bbl/philadelphianeurodevelopmentalcohort.html

n~9500 individuals (age 8-21)
Diverse medical conditions
Comprehensive assessments

1445 participants with MR (resting state, working memory task, ASL, T1w, DTI) - One Site

Genomics (through dbGaP)

Philadelphia Neurodevelopmental Cohort

22 of 78

Human Connectome Project

http://www.humanconnectome.org

The Connectome Coordination Facility (CCF) houses and distributes public research data for a series of studies that focus on the connections within the human brain. These are known as Human Connectome Projects.The CCF currently supports 18 human connectome studies

1200 healthy young adults

There is family structure in the data (MZ and DZ twins)

Genotyping will go up on DB gap

Re-wrote the standard for MR data acquisition, analysis and data sharing

All data (plus preprocessed available through Connectome DB)

Human Connectome Project

23 of 78

Human Connectome Project

24 of 78

https://ndar.nih.gov/

The National Database for Autism Research (NDAR) is an NIH-funded research data repository that aims to accelerate progress in autism spectrum disorders (ASD) research through data sharing, data harmonization, and the reporting of research results.

To request access to detailed human subjects data, you must be sponsored by an NIH recognized institution with a Federalwide Assurance and have a research related need to access NDA data. NDA federated repositories have their own access requirements.

NDAR - the NIMH data archive

25 of 78

https://www.nature.com/articles/sdata2017181

Child Mind Institute

26 of 78

Single subject: Dr. Russel Poldrack
Brain scans twice a week for 18 months
Focused on fMRI data
Key condition: coffee + breakfast
Multi-center (two)
Transcriptomic and metabolomic data

https://github.com/poldrack/myconnectome/

MyConnectome

(Poldrack et al.,2015, Nat Commun.)

27 of 78

(Poldrack et al.,2015, Nat Commun.)

28 of 78

+ many other variables like sleep and surveys

(Poldrack et al.,2015, Nat Commun.)

30 of 78

PhysioNet

https://physionet.org

1980: MIT-BIH Arrhythima Database

1999: 11 CD-ROMs

2016: Laufman-Greatbatch prize

Goldberger et al. “PhysioBank, PhysioToolkit, and PhysioNet: Components of a New Research Resource for Complex Physiologic Signals.” Circulation. 2000

31 of 78

Major resources and platforms

Major Population Cohorts

33 of 78

All of Us

https://www.researchallofus.org/

35 of 78

CanPath (Project for Tomorrow)

50,000 participants genotyped

(Borugian et al., 2010, CMAJ)

Now +

36 of 78

CanPath (Project for Tomorrow)

37 of 78

https://abcdstudy.org/

Longitudinal Assesment of Developing Children

Human Connetome Project Standard MR Aquisition

10,000 children ages 9-10 to join the study at 20 research sites

ABCD

39 of 78

http://www.ukbiobank.ac.uk/

UK Biobank ...is following the health and well-being of 500,000 volunteer participants and provides health information, which does not identify them, to approved researchers in the UK and overseas, from academia and industry.

Plus every other clinical variable + genomics

UK Biobank

40 of 78

UK Biobank

2021 -

42 of 78

Data harmonization - Maelstrom

Unification

Group similar variable types

Map variable categories to a single common set of definitions

Meta-data based

Harmonization

Match two variables directly between two (or more) separate studies

Often requires data transformation (Maelstrom schema)

Time intensive, case-by-case (e.g. weight in lbs vs. kg)

https://www.maelstrom-research.org/

43 of 78

TAY Cohort Study

taycohort.ca

n=3,000

5 years

Ages 11-24

CAMH patients

44 of 78

TAY Cohort Study

Developmental trajectories of psychosis spectrum symptoms

Trajectories of cognitive performance and education

Brain circuits and molecular profiles

Service use patterns and health care costs

Trajectories of whole-person networks

45 of 78

Major resources and platforms

CAMH BrainHealth Databank

46 of 78

KCNI Summer School July 12, 2021

Joanna Yu, PhD

48 of 78

CAMH’s Achievements Towards Learning Health System

49 of 78

BrainHealth Databank

Considerations for Developing AI/ML/Modeling Tools & Technology to Advance Personalized Treatment

Sustained collection of high quality standardized clinical data

Essential for establishment of reliable clinical decision support

Accumulation of data reflective of ‘real-world’ patient population

Facilitate translation back into clinical practice
Consider caveats, i.e. representativeness

Access to ‘big data’ – rich longitudinal, multiscale treatment patient trajectories – that can be reused to answer many questions

Requires robust data management to ensure longevity of data

50 of 78

BrainHealth Databank Key Objectives

New Technology Infrastructure:

51 of 78

Digital Measurement-Based Care

52 of 78

2. Integration of Research Measures into Care Pathways

53 of 78

3. AI & Data Driven Personalized Care

Clinician Clinical Decision Support Dashboard for Major Depressive Disorder Integrated Care Pathway (MDD-ICP)

54 of 78

Developing New Tools - Artificial Intelligence & Machine Learning - to Advance Clinical Decision Support

Considerations: Acknowledgment of potential limitations of AI/ML tools as appropriate, i.e. diversity of sample

55 of 78

3. AI & Data Driven Personalized Care

56 of 78

4. Open Science for Discovery & Innovation

57 of 78

CAMH MDD-Integrated Care Pathway Patient Trajectories

58 of 78

REDCap

LabKey

XNAT

Cerner

Nexus

Knowledge Graph

Cohort Explorer

Influx

BrainHealth Databank Research Resource

59 of 78

FINDABLE

Cohort Explorer: Interactive dashboard to search metadata enabling discovery of datasets to support hypothesis and data driven analyses.

ACCESSIBLE

Data Steward: Resource to support generation of datasets. Access granted with appropriate approvals.

INTEROPERABLE

Adoption of Standard Ontologies: FHIR (Fast Healthcare Interoperable Resources) is international standard ontology for representing EHR data with flexible integration with other data types. Data models represented in searchable knowledge graph to connect and show relationships between data, i.e. patient’s care pathway.

REUSABLE

Standardized Data Documentation: Detailed description of data provenance, i.e. who collected the data, when it was collected, in what context, protocol changes, etc., to ensure and allow data to be reused for years to come

Adoption of FAIR principles

60 of 78

Continuous Learning Health System

ainHealth Data Bank Framework

61 of 78

CAMH-Wide BrainHealth Databank Team

62 of 78

ready to analyze

but not sure where to find relevant data

“Digital Health Data ”

66 of 78

Key Questions

(before & after you have data)

Raw sensor-level data is NOT enough

Key unit of analysis and data linkage

Individual
Socio-demog factors
External factors

Environment
Neighborhood characteristics

..many more ...

https://www.ukbiobank.ac.uk/enable-your-research

67 of 78

Understanding context around data

From whom
When
How long
Using what mechanism
Under what conditions

71 of 78

Demo 1

Demo 2

73 of 78

More examples

74 of 78

public health data

fingertipsR

https://www.ebi.ac.uk/about/news/feature-story/biobanks-genetic-data-demand

75 of 78

public health data

fingertipsR

https://www.ebi.ac.uk/about/news/feature-story/biobanks-genetic-data-demand

76 of 78

dhealth.synapse.org

78 of 78

Session 3: In case you want to follow along

Create an account on synapse - www.synapse.org
If you want to upload data - Get certified - https://www.synapse.org/#!Quiz:Certification

1 of 78

2 of 78

3 of 78

4 of 78

5 of 78

6 of 78

7 of 78

8 of 78

9 of 78

10 of 78

11 of 78

12 of 78

13 of 78

14 of 78

15 of 78

16 of 78

17 of 78

18 of 78

19 of 78

20 of 78

21 of 78

22 of 78

23 of 78

24 of 78

25 of 78

26 of 78

27 of 78

28 of 78

29 of 78

30 of 78

31 of 78

32 of 78

33 of 78

34 of 78

35 of 78

36 of 78

37 of 78

38 of 78

39 of 78

40 of 78

41 of 78

42 of 78

43 of 78

44 of 78

45 of 78

46 of 78

47 of 78

48 of 78

49 of 78

50 of 78

51 of 78

52 of 78

53 of 78

54 of 78

55 of 78

56 of 78

57 of 78

58 of 78

59 of 78

60 of 78

61 of 78

62 of 78

63 of 78

64 of 78

65 of 78

66 of 78

67 of 78

68 of 78

69 of 78

70 of 78

71 of 78

72 of 78

73 of 78

74 of 78

75 of 78

76 of 78

77 of 78

78 of 78