cceHUB
A Knowledge Discovery Environment
for Cancer Care Engineering Research
Ann Christine Catlin
HUBzero Workshop
November 7, 2008
a HUB for Cancer Care Engineering
integrate/synthesize
biological OMIC data
 biomarker knowledge
• OMIC Analysis Labs
• Statistical Modelers
• Visual Analytics
HUB
A series of somatic mutations leads to clonal
expansion and an adenoma. Mutation of an
Support for
Community-shared
Resources
Click to Next Slide
CCE integrated project hierarchy
The Cancer Care Engineering (CCE) project is a highly innovative,
interdisciplinary, multi-institutional endeavor that holds promise for
revolutionizing the current paradigms of cancer prevention, detection,
treatment and care delivery by focusing on translating cancer research into
clinical practice …
The current challenge is to focus the rapid and extraordinary advances using
large scale human OMIC analyses to significantly improve survival rates in
community-based oncology clinics and develop community-wide prevention
efforts. In fact the tremendous scientific advances and rapidly developing
capabilities offer an enormous opportunity to catalyze improvement by
focusing on the engineering of cancer care. Our vision addresses this
opportunity by applying systems engineering principles and unique data
visualization and statistical model building to the broad spectrum of cancer
prevention, treatment, and care delivery. CCE focuses on identifying
opportunities for improvement by analyzing system behavior against best
practices and creating predictive models for the treatment and care of
patients. Moreover, cancer care delivery is included as part of the cancer
treatment system, where novel statistical models are used to predict both
disease behavior (effective treatment decisions) and system response
(efficient cancer care delivery).
Cancer Care Engineering Projects
funded by the Regenstrief Foundation
CCE-1: A Multi-Agent Approach to Modeling
of the Indiana CRC Care System
CCE-2: An Indianapolis CRC Quality
Improvement Initiative
…
CCE-5: A Fusion Center for Cancer Care
System Information – the Cancer Care
Situation Room
… highly innovative, interdisciplinary, multi-institutional endeavor …
CCE White Paper, 2007
CCE-6: Information Infrastructure and Raw
CCE TEAM: more than 70 scientists, clinicians,
physicians, nurses,
Data statisticians,
Analysis
Indiana University Simon Cancer Center
Purdue Cancer Center
Oncological Sciences Center
e-Enterprise Center
Regenstrief Center for Health Care Engineering
Regenstrief Institute
Roudebush VA Center for Implementing Evidence-Based Practice
IU Center for Health Services and Outcomes Research
Department of Medicine, IU School of Medicine
Department of Medicinal Chemistry and Pharmacology, Purdue University
School of Chemical Engineering, Purdue University
School of Electrical and Computer Engineering, Purdue University
Department of Statistics, Purdue University
engineers, computer scientists, health service
researchers, university and hospital
CCE-7: Augmenting Physical Sample
staff
Collection, Clinical Data Collection, OMIC
Laboratory Analysis, Conversion to Digital
Data
…
Click to Next Slide
CCE team: collaboration !
Integrated Hierarchy of Projects
CCE-1: A Multi-Agent Approach to
Modeling of the Indiana CRC Care
System
CCE-2: An Indianapolis CRC Quality
Improvement Initiative
…
CCE-5: A Fusion Center for Cancer
Care System Information – the Cancer
Care Situation Room
CCE-6: Information Infrastructure OMIC Data Collection and Analysis
CCE-7: Augmenting Physical Sample
Collection, Clinical Data Collection,
OMIC Laboratory Analysis, Conversion
to Digital Data
…
Click to Next Slide
Using HUB Technology I
• Content Management System for Scientists
• Collaboration and Social Networking
Click to Next Slide
Integrative Mathematical Models
Click to Next Slide
Shared models need shared data …
Click to Next Slide
Using HUB Technology II
• Unique Middleware for Modeling and Simulation
Rappture
Physical Machine
modeling
code
collaborator
Content
Database
Maxwell’s
Daemon
Middleware
Virtual Machine
tool session cluster
VIOLIN
scientist
rendering farm
Click to Next Slide
visualization servers
What we need to support that’s new …
 OMIC workflow
 biosamples  biological data  biomarker knowledge …
“data lifecycle” support
 end-to-end user support
 Data … the new shared resource
 data repository & data support infrastructure
 metadata … annotate, track, characterize content
Click to Next Slide
OMIC experiment workflows
IU Simon
Cancer Center
Sample acquisition
processing
transfer
transfer to Purdue
sample
Bindley
Biosciences
Laboratories
LECO
GC GC MS
instrument generated
Sample preparation
Instrument analysis
Data generation
raw instrument data
pre-processed data
RAW
DATA
clinical database,
demographics,
diet, diagnosis,
treatment
BLOOD SAMPLE
CRC PATIENT/CONTROL
35MB
peak list
TXT
protocols
Document library
& document tagging
tracking
Metadata processing
external DB
linkage
Communication &
information exchange
annotation
data
converters
CDF
1.2GB
retention times,
m/z, intensities
data
converters
PREPROCESSED
DATA
Interactive GUI
350MB
spectrum, list of peaks,
PEG
CSV
annotation
DAT/RAW
SAMPLE DATA
converted, reduced,
selected, filtered,
transformed
more ...
datasets
methods
upload
Interactive GUI
Document library
& document tagging
Data transfer
tracking
Metadata processing
storage
File server & backup
metadata Metadata processing
linkage
Click to Next Slide
cceHUB
OMIC data analysis workflows
Laboratory Analysis
Statistical Modeling
Visual Analytics
Processing multiple samples
Comparisons across samples
Modeling across samples
Visualization across samples
data exploration
sample
sample
sample
sample
sample
ontology support
sample
ANALYZED
DATA
datasets
visualization
modeling
data tracking
tool tracking
data capture
annotation
metadata linkage
storage
Click to Next Slide
cceHUB
a HUB data support framework
Data Explorer
Metadata-based Query
Data Upload
Process Tracking
Results Capture
Data Annotation, Content Characterization & Tracking
OMICS Data File Repository
Metadata Database
Click to End