Informatics Tools for Molecular-Based Specimen Banks

1
Informatics Tools for Molecular-Based Specimen
Banks

• Current Resources
• CaTIS Database
• VGSR Concept

Mark A. Watson, M.D., Ph.D.
2
Informatics Limitations of Specimen Banks

• Inability to completely and accurately annotate
  specimens with clinical and pathological data.
• Inefficient tracking of specimens and specimen
  quality to multiple different research projects.
• Inability to perform real-time queries of
  readily available samples from secure data
  servers.
• Lack of biological (research) data annotation
  for specimens.

• Inability to completely and accurately annotate
  specimens with clinical and pathological data.
• Inefficient tracking of specimens and specimen
  quality to multiple different research projects.
• Inability to perform real-time queries of
  readily available samples from secure data
  servers.
• Lack of biological (research) data annotation
  for specimens.

3
SCC Tissue Procurement Core Overview

• Established 1997
• 24 studies
• Archival bank
• Institutional studies
• American College of Surgeons Oncology Group
  (ACOSOG)
• 21,000 specimens / 8,000 patients (sporadic
  cancer)
• Frozen Tissue / Paraffin Blocks / Serum
• 6,300 DNA and RNA samples
• 112 Sample distributions
• Intra- and extramural investigators

4
SCC Tissue Procurement Bioinformatics

• Mark A. Watson, M.D., Ph.D.
• Dir. Siteman Cancer Center Tissue Procurement
  Core Facility
• Dir. ACOSOG Central Specimen Bank
• Dir. Siteman Cancer Center Multiplexed Gene
  Analysis Facility
• Rakesh Nagarajan, M.D., Ph.D.
• Asst. Dir. Siteman Cancer Center Bioinformatics
  Core Facility
• Jeff Milbrandt, M.D., Ph.D.
• Dir. Siteman Cancer Center Bioinformatics Core
  Facility
• Richard Wilson, Ph.D.
• Dir. Washington Univ. Genome Sequencing Center
• (5) Laboratory FTEs
• Persistent Systems (Software Development)

5
SCC Tissue Procurement Core Capabilities

• Specimen Collection
• Multiple institutional sites
• Outside institutions
• Specimen Storage
• LN2, -80C, 4C
• Specimen Processing and QA
• DNA and RNA samples
• Sample Arraying
• Tissue / DNA / RNA
• Data Management (Tracking)

6
A Paradigm for Specimen Data Utilization
Specimen Data
Multi-Dimensional Data Space
7
CaTIS Specimen Database

• MS Access Based
• Non-networked
• Scalable
• Rapid customization
• Functionality
• Patient / Specimen / Sample Accession
• Storage / QA data (tissues and samples)
• Pathology data (manual entry)
• Distribution data
• Mapping to sample arrays / robotics
• Tracking through to outside institutions (e.g.
  WU-GSC)
• Mapping to experimental results

8
CaTIS Specimen Database
9
CaTIS Specimen Database
10
CaTIS Specimen Database
11
CaTIS Specimen Database
12
CaTIS Specimen Database
13
CaTIS Specimen Database
14
CaTIS Specimen Database
Specimen Data Patient Code Number Specimen Code
Number Specimen Info / QC Submission Data Path
Data
Sample Data Specimen Code Number Sample Code
Number Sample QC Investigator Distribution Experim
ental Results
Patient Data Patient Code Number Demographics Stud
y / IRB
15
Aim 1 Commercialize and Distribute Key CaTIS
Components

• Data security scheme
• Migrate to web based data entry / query
• Migrate to platform independence
• Open interface to other pathology data systems
• Open interface to genomics data systems (e.g.
  CHIPDB)
• Use of caBIG-defined common data elements
• Pathology
• Molecular
• Additional tools (e.g. pedigrees) suggested by
  adopters

16
Virtual Genomic Sample Repository (VGSR)

• Most specimen resources do not allow direct
  querying to sample detail.
• Most specimen resources do not allow querying by
  biological (experimental) data.
• Caucasian male / gt 60 YO / T2N0 NSCLC
• T2N0 NSCLC / p53- / 50 ug DNA / U133 Array
  4567
• Single specimens may be used for multiple
  studies at the genome / transcriptome / proteome
  level
• Maximum specimen utility
• Integrative systems biology

17
Virtual Genomic Sample Repository (VGSR)
Goal To develop an informatics system and
scientific culture to facilitate the sharing of
molecular biospecimens and biological data
associated with their use.
18
VGSR Conceptual Data Flow
Institutional Tissue Banks
19
VGSR - Components

• Policy / Governance
• Can samples be shared (IRB / MTA) ?
• Will samples be shared (ownership / IP) ?
• How will samples be shared (review /
  prioritization) ?
• Sample QA / Data standards
• Central Database Server
• Web-based Applications
• Sample Registration
• Sample Query / Request
• Sample Manager
• Data Registration

20
VGSR Key Features

• No other system currently meets this need
• Cooperative Group Tissue Banks
• Pathology Database Tools
• Unique opportunity to test the Virtual Specimen
  Bank concept
• Numerous points of interoperability with caBIG
  systems
• Based on caBIG Architecture Workspace
  recommendations
• Use of caBIG vocabularies and common data
  elements
• Integrated with NCICB tools (caIMAGE)
• Works with other software developed in
  Integrative Cancer Research Tools Workspace (e.g.
  Function Express, Mutation Viewer)

21
CaTIS / VGSR Required Resources

• Developmental Requirements
• Consensus building / SOPs / CDEs
• Personnel
• 1 Architect (Shared with Architecture Workspace)
• 1 Project manager (50- Shared with Integrated
  Cancer Research Tools Workspace)
• 1 DBA (50- Shared with Integrated Cancer
  Research Tools Workspace)
• 2 Programmers
• Hardware
• Enterprise-class server / programmer
  workstations
• Software
• Rational Rose/XDE, ClearCase (CVS), ClearQuest
• JBuilder, C Builder, Ant
• RDBMS (e.g. Oracle, DB2, MSSQL, etc.)

22
CaTIS Project Timeline

• Month 1-6
• Conversion of CaTIS to Commercialized CaTIS
• Month 7-12
• Deploy CaTIS API to extramural sites (adopters)

23
VGSR Project Timeline (1-12 months)

• Month 1-8
• Consensus / Policy building
• Architecture design / CDEs and vocabularies
• Month 9-12
• Database building
• Web tools design

24
CaTIS / VGSR Outcome Measures

• Utilization of CaTIS by adopter sites
• Successful registration of samples to VGSR
• Utilization of VGSR samples by new collaborative
  research teams for translational cancer research
• Funding opportunities
• Publications track record