Efficient genomic profiling of patients: the benefit of systems interoperability

Abstract

Genedata ProfilerTM is a translational research software platform developed in collaboration with leading
pharmaceutical companies to effectively process, manage, and analyze omic and phenotypic data to the
highest standards of data quality and regulatory compliance. Genedata Profiler complements the knowledge
management platform tranSMART by standardizing the processing and quality control of omics data,
simplifying the publishing of data into the data warehouse, and adding sophisticated statistical analyses. In
this case study we will demonstrate an end-to-end workflow to identify insulin response biomarkers utilizing
the seamless, bidirectional interoperability of Genedata Profiler with tranSMART

Full text

Efficient genomic profiling of patients:
the benefit of systems interoperability
Improving the process of translational R&D
Achieving the vision of precision medicine is reliant
to a large extent on translational research activities.
Such activities require researchers to characterize
and profile patients using omic technologies in order
to understand their response to new therapies,
stratify patients for trials or search for new disease
biomarkers
1
. Genedata Profiler is an enterprise
software platform used by pharma and biopharma
companies to address these challenges and to
optimize the process of translational research.
In this case study, we will discuss how Genedata
Profiler complements existing in-house software
solutions such as tranSMART
2
to create a complete
infrastructure for performing translational research.
The open-source data warehouse tranSMART is used
by several pharmaceutical companies to manage multi-
omics data sets to be used for biomarker projects.
Often, tranSMART is used to merge ‘clinical data’ from
the highly regulated clinical study environment into
the less-regulated ‘discovery’ research environment,
where users have much more flexibility to add omics
data as well as other important research information.
By integrating Genedata Profiler with tranSMART,
we have enhanced and expanded the capabilities of
organizations to perform patient genomic profiling.
The result is a very powerful, user-friendly platform
for translational research.
Axel Schumacher a, Mark A. Collinsb, Marc Flescha, Miles Fisher-Pollardc, & Tamas Rujanc
aGenedata GmbH, Munich, Germany | bGenedata Inc, Lexington, USA | cGenedata AG, Basel, Switzerland
Keywords: tranSMART, quality control, data harmonization, translational research, precision medicine, patient privacy
#2
Fig. 1: Left: Version-controlled workflows for repeatable execution. The workflow engines of Genedata Profiler allow organizations to build
and deploy standardized workflows throughout an organization, ensuring data quality and reproducibility. Example shows an RNA-Seq
pipeline that was ‘approved’ by the study manager. Right: The best-in-class, fully interactive genome browser incorporated into Genedata
Profiler facilitates visual inspection of raw data such as original short read alignments together with analysis to ensure result consistency.
Genedata Profiler
TM
is a translational research software platform developed in collaboration with leading
pharmaceutical companies to effectively process, manage, and analyze omic and phenotypic data to the
highest standards of data quality and regulatory compliance. Genedata Profiler complements the knowl-
edge management platform tranSMART by standardizing the processing and quality control of omics data,
simplifying the publishing of data into the data warehouse, and adding sophisticated statistical analyses. In
this case study we will demonstrate an end-to-end workflow to identify insulin response biomarkers utiliz-
ing the seamless, bidirectional interoperability of Genedata Profiler with tranSMART.

Schumacher et al. Efficient genomic profiling of patients: the benefit of systems interoperability
2
published with a click of a button in minutes rather
than hours (Fig. 3).
In addition, Genedata Profiler allows augmentation
of studies without the need to reload all data for
each study. Out-of-the-box integration with public
omics data sources such as GEO, ArrayExpress,
and various other resources, allows researchers
to integrate data from various public repositories
easily into tranSMART. The easy access to a wide
variety of multi-omics data enables efficient data
discovery and data sharing.
Harmonizing raw omics data processing
& quality control
Data quality and hence data curation is critical
to making the right scientific conclusions.
Genedata Profiler complements tranSMART by
adding sophisticated data processing & curation
capabilities to harmonize and standardize data
processing workflows. Expert users can set up and
approve such workflows and make them available
to a larger user community (Fig. 1).
To illustrate the patient profiling capabilities of
Genedata Profiler and the value of the integration
with tranSMART, we sought to identify biomarkers
for insulin response in skeletal muscle. We applied
a best-practice workflow (Fig. 1, left) to process
NGS and microarray data from a public RNA-
Seq data set of human skeletal muscle myocyte
samples
3
( n = 1 2 , 3 d i f f e r e n t t r e a t m e n t t i m e p o i n t s )
together with expression profiles from a different
set of muscle biopsies (n=36), which were analyzed
with Affymetrix microarrays
4
.
Inspection of the comprehensive quality control
report (Fig. 2) automatically generated by the
workflow indicates that all omics data is of sufficient
quality to be utilized for statistical analysis, and is
ready to be published to tranSMART.
Simplifying data loading into tranSMART
tranSMART utilizes a highly complex Extract-
Transform-Load (ETL) infrastructure to allow
expert users to load data into tranSMART. Loading
omics and clinical sample annotations (metadata)
into tranSMART can therefore be time consuming
and expensive.
Genedata Profiler uses its own public APIs to load
data directly into tranSMART, allowing data to be
Fig. 2: Automated quality control reporting.
Configurable and
dynamic quality reports in Genedata Profiler provide a rich set of
quality metrics on NGS reads.
Fig. 3: Left: Saving high-dimensional data to tranSMART. User selects a study from the data warehouse study tree together with other data
groups to be uploaded. The data is uploaded, processed and written to tranSMART and the user notified when data is available. Right: Saving
clinical data to tranSMART. Additional clinical annotations can be selected from a list by dragging and dropping. Omic and clinical annota-
tions may be linked by choosing an appropriate field, e.g. subject ID.

Schumacher et al. Efficient genomic profiling of patients: the benefit of systems interoperability
3
Fig. 4: Left: Initiating data transfer from tranSMART to Genedata Profiler.
After selecting high- and low-dimensional data nodes in
tranSMART, data is automatically transferred to the statistical module of Genedata Profiler. Middle: The Volcano Plot visualizer (top) displays
a scatter plot of markers in which P-values are plotted against n-fold change. The most significant markers (red) are highly expressed after
insulin treatment. A gene ontology Fisher’s Exact Test (bottom) indicates that these genes are involved in cellular responses to zinc ions, which
play a central role in glycemic control. Right: The Genedata Profiler platform was specifically tailored for the integration & interpretation of
experimental data in translational R&D, providing a wide range of data analyses through a rich statistical toolbox and intuitive visualizations.
Identifying & reporting clinically relevant
biomarkers
Genedata Profiler makes tranSMART smarter
by adding sophisticated statistical methods not
available in tranSMART. It offers:
• A rich statistical toolbox to perform a wide range
of data analyses;
• External algorithms as plugins;
• Integration of data across technologies and
studies from in-house and public data sources;
• Sophisticated data visualization and reports.
Multi-omics approaches inherently increase the
already growing complexity of data in the life
sciences. Genedata Profiler provides scalability
beyond the capabilities of tranSMART so that huge
amounts of complex data can be analyzed within a
short time.
Using the statistical tools of Genedata Profiler
(Fig. 4), we were able to identify a novel biomarker
Gadd45A
(a diabetes-associated gene), potentially
linking diabetic cardiomyopathy and baroreflex
dysfunction, which was not detected in the original
study by Coletta et al
4
.
Making the data regulatory-compliant
As pharmaceutical companies increasingly leverage
confidential patient data such as medical records
and genetic information in translational research,
important regulations governing data security and
privacy must be respected. Recent high-profile
decisions, e.g. replacement of the EU-US Safe Harbor
agreement on transatlantic data exchange with a
new “Privacy Shield”, impose stronger compliance
requirements on organizations. Genedata Profiler
provides comprehensive capabilities to ensure
patient privacy and maintain the chain of custody of
data, goals that are core to regulatory compliance.
For example, in our biomarker case study, the
integration with tranSMART ensures that the same
access controls are used between Genedata Profiler
and tranSMART, safeguarding the insulin study
patient data and reducing exposure to regulatory
non-compliance risk.
The information infrastructure of Genedata Profiler
enables organizations to cope with the wide range
and volumes of omics data as well as the wide
variety of data consumers (e.g. bioinformaticians,
analysts, biologists, etc.).

Genedata ProfilerTM is part of the Genedata portfolio of advanced software solutions that ser ve
the evolving needs of drug discovery, industrial biotechnology, and other life sciences.
Basel | Boston | Munich | San Francisco | Toky o
www.genedata.com/profiler | profiler@genedata.com
© 2016 Genedata AG. All rights reserved. Genedata Profiler is a registered tr ademark of Genedata AG. All other pro duct
and ser vice names mentioned are th e trademarks of their re spective companies.
Schumacher et al. Efficient genomic profiling of patients: the benefit of systems interoperability
Strict user role management, audit trails, access
authorization, data federation, data lifecycle
and method management, and a comprehensive
reporting infrastructure are core components of
the software which enable:
• Collaboration between different user roles
throughout a global organization using
comprehensive role-based access controls;
• Integration, federation, and curation of the wide
variety of omic, phenotypic and patient data from
internal (e.g. tranSMART), external and public
data sources;
• Sharing of data, methods and results.
Summary
Genedata Profiler makes omic-based patient
profiling processes significantly more efficient. It
streamlines the whole data processing, analysis,
and management process, and reduces the time
it takes to perform biomarker studies. As we have
shown in this case study, Genedata Profiler and its
bidirectional integration with tranSMART can help
scientists gain new insights into omics data from
clinical studies, while reducing compliance risk in
their translational research.
References
1. Schumacher, A., Rujan, T. & Hoefkens, J. A collaborative approach
to develop a multi-omics data analytics platform for translational
research. Appl. Transl. Genomics 4–7 (2014). doi:10.1016/j.
atg.2014.09.010.
2. Athey, B. D., Braxenthaler, M., Haas, M. & Guo, Y. tranSMART: An
Open Source and Community-Driven Informatics and Data Sharing
Platform for Clinical and Tr anslational Research. AMI A Jt. Sum mits
Transl. Sci. Proc. AMIA Summit Transl. Sci. 2013, 6–8 (2013).
3. Väremo, L. et al. Proteome- and Transcriptome-Driven
Reconstruction of the Human Myocyte Metabolic Network and Its
Use for Id enti cation of Mar kers for Diabe tes. Cell Rep. 11, 921–933
(2015).
4. Coletta, D. K. et al. Effect of acute physiological hyperinsulinemia
on gene expression in human skeletal muscle in vivo. Am.J.Physiol
Endocrinol.Metab 294, E910–E917 (2008).
Please cite this article as:
Schumacher, A., Collins, M., Fisher-Pollard, M. & Rujan,
T. (2016) Efcient genomic proling of patients: the benet of
systems interoperability, doi: 10.13140/RG.2.1.3093.4648
Fig. 5: Left: Study-centric, role-based web UI.
The web interface facilitates collaboration, data, method and results sharing. The
underlying enterprise architecture provides integration/federation with HPC file systems, internal and external databases and public data
sources. Right: User role management in Genedata Profiler. The ‘Members’ dialog lists all the members of a selected study and the specific
roles that are associated with those members. Roles are used to define permissions and access control.