Track 9 - April 5 – 7, 2016
Pharmaceutical R&D Informatics
Collaboration, Data Science and Biologics
With the increased generation of data from a wide number of sources including R&D, clinical, translational, genomic and personal data, pharma and biotech must effectively manage and integrate data from all stages of the pharmaceutical value chain to enable more informed decisions. Track 9 explores the transformation of current IT and informatics teams into data science groups and current progress made by such groups in the analysis, integration and visualization of complex data sets, including genomic, imaging, clinical, external/internal collaboration and real world data.
Tuesday, April 5
8:00 – 11:30 Recommended Morning Pre-Conference Workshops*
Data Management for Biologics: Registration and Beyond
12:30 – 4:00 pm Recommended Afternoon Pre-Conference Workshops*
Data Science Driving Better Informed Decisions
* Separate registration required
7:00 am Workshop Registration and Morning Coffee
2:00 – 6:00 Main Conference
5:00 – 7:00 Welcome Reception in the Exhibit Hall with Poster
Wednesday, April 6
7:00 am Registration Open and
9:00 Benjamin Franklin Awards and Laureate Presentation
9:30 Best Practices Awards Program
9:45 Coffee Break in the Exhibit Hall with Poster Viewing
10:50 Chairperson’s Opening Remarks
Yuriy Gankin, Ph.D., Chief Life Sciences Officer; Vice
President, Life Sciences, EPAM Systems, Inc.
11:00 Leverage the Wealth of Internal and External Information to Drive Collaboration and Project-Centricity into Your Research Informatics Platform for Drug Discovery and Development; A Strategic Imperative
James Connelly Ph.D., Global Head, Research Data Management,
Every pharmaceutical company holds a wealth of information collected for numerous drug candidates during research, pre-clinical and clinical development and for marketed drugs. Much of this information is trapped, undiscoverable and not optimally useable for drug discovery and development. Also, it cannot be effectively combined with the massive amount of externally available data. Two pilot studies with IBM Watson successfully extracted high quality information from toxicology reports and enabled researchers to discover critical insight for drug-repurposing proposals. Cloud-based SAR data platforms will further stimulate the use of broad data sources for research and create a project-centric and collaborative environment. Increasing reliance of Pharma on a diverse range of drug discovery collaborations has caused severe challenges to adapt Big Pharma datawarehouse-centric SAR data platforms. Evolution and innovation of cloud-based SAR data platforms has revolutionized our approach to SAR data integration, sharing and analysis. We will describe our current approach to collaboration data and the evolutive roadmap towards a fully capable SAR data platform in the cloud that will utilize cloud-based services and big data technology. This will reduce the cost of internally supported systems and create a scalable external SAR informatics system. These are major trends in the pharma industry and are driven by the acceptance, evolution and scalability of cloud-based and computational services and for collaborative SAR data sharing and analysis. There is a strong value-added in utilization of this wealth of information for R&D and strategic imperative to build informatics systems using Big Data technology to access and utilize all relevant information sources.
11:30 Making Scientific Data 100x Easier to Use: Transforming
Pharmaceutical R&D with Scalable Approaches to Data Stewardship and Data
Carol Rohl, Ph.D., Executive Director, Scientific Information
Management, Merck & Co., Inc.
Arguably, the vast majority of time and resources in all scientific analytics and informatics projects are dedicated to finding, accessing, understanding, curating, and integrating the input data assets. While scientific data is generally effectively managed for its primary use, it often lacks the context that facilitates secondary uses and cross-functional integration. As a result, much of the research informatics efforts across pharma are focused on creating solutions to these challenges for specific sets of use cases within a particular problem space. As the use of predictive modeling and analytics increases to address the challenges of declining R&D productivity and increasing pressures for demonstrating product value, scalable approaches are required to handle the ever increasing variety of data types, data sources, data models and analytics patterns. To address these challenges at scale, we are combining data stewardship tools and capabilities that leverage crowdsourcing across the community of data creators along with a platform approach to solving data variety problems, building on the Big Data technology stack, which enables an ecosystem of agile fit for purpose datasets and informatics solutions.
12:00 pm Informatics Solutions to Address
the Changing Research Paradigms
Robert Brown, Ph.D., Vice President, Global Informatics,
Conducting research projects across multiple organizations presents a number of challenges which must be overcome for them to be successful. Using case studies from industry and academia, this talk will discuss how dedicated hosted informatics systems designed to support collaborative small molecule and biologics research can help enhance the success of these projects.
12:15 Integration of Rich, Connected Analytical Information Across Corporate Informatics Landscapes and the Impact on Innovation
Andrew Anderson, Vice President, Global Informatics, Advanced Chemistry Development, Inc.
A variety of informatics tools are utilized to facilitate lab-to-lab, or scientist-to-scientist collaboration. However, these tools often require “abstraction” of certain data types – especially data resulting from analytical experiments conducted to characterize materials. This case study aims to provide a overview of how forward-looking organizations are enriching their collaboration interfaces with rich, standardized, live data “assemblies."
12:30 Session Break
12:40 Luncheon Presentation II: If 80% of Your R&D Was Externalized, Could Your Current Informatics Infrastructure Support it?
Scott Weiss, Ph.D., Senior Director, Product Management & Product Strategy, IDBS
Externalization has fundamentally changed the way R&D teams plan work, share information and secure IP assets in a dynamic environment of ever changing partners. This presentation will review IDBS’ E-WorkBook Connect a new, Cloud-based Informatics Solution, designed specifically to provide an easy to use, secure environment for B2B collaboration.With Connect, business users can quickly set up new projects, invite teams, capture and review content and securely publish data, document content back to the corporate systems.
1:10 Session Break
1:50 Chairperson’s Remarks
Peter A. Covitz, Ph.D., Senior Director, Research and Translational IT, Biogen
1:55 Improving Decisions with Richer Data: Text Mining and
James Gill, Ph.D., Director, Research Analytics, Bristol-Myers
The quality of decisions are based on a number of factors, not the least of which is the quality of the supporting data. Recently there has been much ado about the amount of untapped data living in unstructured formats primarily in text documents and how it can be used. What makes text particularly valuable is the amount of supporting context often available. Imagine you are interested in the potency of a drug. Querying the traditional data warehouse might return a result like Assay1, IC50=3.25uM. However a power point slide or word document might say “when tested in an in vitro cellular assay IC50 values around 3.25uM were observed however this compound has significant solubility issues”. Clearly the latter provides a richer more informative perspective. Of course the trick is finding the entities in the text and then formatting the results so they are useful for the decision makers. Our group has been focusing on provisioning search of unstructured content for a number of years. We will discuss how we use different technologies in various combinations to provide data to BMS in a sustainable way by delivering services along a spectrum of solutions including self-service tools for recurring questions (current awareness, gene/indication relationships), advanced analyst driven research aimed at specific topics (drug repositioning, TR&D research, target ID) and mission critical searches requiring specialized analysts (Patent and specialized literature searching). In particular in this talk we will dive into ongoing work to improve the findability of documents via text search and the utility of text mining across our portfolio.
2:25 Of Soft Drinks and Software: A Framework to Sustain
Innovation in Computational Molecular Design
Enoch Huang, Ph.D., Executive Director, Computational
Sciences, Pfizer R&D
One of the challenges in sustaining innovation in
computational molecular design is the need to harness and deliver promising
solutions irrespective of their source, without creating new software
applications or re-write existing ones. In this talk, I will provide an
overview of Pfizer’s molecular design infrastructure and the paradoxical
solution that we’ve used to address this classic dilemma. I will also provide
examples of specific computational methods and algorithms enabled by this
approach, and describe our recent foray into cloud computing.
2:55 How Smart Data Transforms Life Science R&D
Jim LaPointe, Managing Director, Pharma & Life Sciences, Cambridge Semantics
Learn how Smart Data solutions are transforming the R&D landscape through better Competitive Intelligence, Site Intelligence & Selection, Clinical Trial Data Integration & Discovery, Scientific Data Integration & Collaboration, PharmacoVigilence & Safety Surveillance, and Real-World Evidence Driven Clinical Trial Design. Cambridge Semantics is putting Big Data analytics into the hands of R&D teams for immediate data insights and business value.
3:10 Taking Scientific Collaboration in the Cloud to the Next Level
Ton van Daelen, ScienceCloud Product Director, Marketing,
Dassault Systemes, BIOVIA
Externalized collaborative research projects require
integration and analysis of compound and bioactivity data from multiple
sources. ScienceCloud supports secure data sharing in a “Cloud” and facilitates
pipelining of data to/from internal data systems. In addition, ScienceCloud is
now also used by small and medium research organizations to serve as their
primary compute infrastructure, drastically lowering TCO while increasing
3:25 Refreshment Break in the Exhibit Hall with Poster Viewing
4:00 Impacting Clinical Research with Data Analytics
James Cai, Head Data Science, Pharmaceutical Research and
Early Development Informatics, Roche Innovation Center New York
Data come in all shapes, sizes, and colors in today’s
clinical research and development. Whether single-protein biomarker data or
high dimensional sequencing data, structured or unstructured, numeric or in
pixels, we need to understand them and manage them effectively. In this
presentation, I will describe several new challenges facing clinical research
teams, and how Data Science has been critical in addressing them. I will
provide examples of how data visualization, machine learning, and image
analysis all contribute to uncovering new insights that impacted scientific or
4:20 From Genome Exploration to Clinical Implementation: The
Challenge of Translational Pharmacogenomic Informatics
Peter A. Covitz, Ph.D., Senior Director, Research and Translational IT, Biogen
The rapidly expanding universe of available human genomic data is finally driving faster and deeper understanding of human genetic variation and how it impacts drug-disease interactions. Genetic biomarker analyses, once an esoteric component of relatively few human drug trials, are now becoming common, and for some programs essential. The trend, driven by better and lower cost nucleic acid sequencing, has created new demands for computational tools and infrastructure. In the realm of discovery science, tools and infrastructure evolve rapidly, keeping pace with the availability of ever-larger data sets and increasingly scalable industrial cloud computing. In the realm of regulated clinical drug trials, however, informatics infrastructure and tooling is – by necessity - far more stable. This presentation will discuss the challenges that arise when these two very different environments intersect, and will describe strategies for addressing the impedance mismatch.
4:40 Building an Operational Data Repository for Exploratory
Al Wang, Associate Director, Exploratory Clinical &
Translational Research IT, Bristol-Myers Squibb
Data integration continues to be a significant impediment to
translational research and development. In particular, the ability to flexibly
combine diverse biomarker data with relevant patient-level information to
produce analysis-ready data sets is a challenge that has not been fully solved
by existing tools and approaches. This talk will describe the use cases around
biomarker data in drug research & development, as well as an ongoing
project to implement a platform that assists in these use cases.
5:00 Accelerating Discovery with Data Integration
Frederik van den Broek, Ph.D., Consultant, R&D Solutions, Elsevier
Deriving insights for R&D decision-making from data requires capabilities and processes to navigate the deluge of data. This talk will present use cases that show once data is correctly aggregated, normalized and integrated, it allows for an integrated view across multiple disparate data sets, such as scientific literature, patents as well as internal databases from in-house researchers and their external collaborators.
5:30 – 6:30 Best of Show Awards Reception in the Exhibit Hall
with Poster Viewing
Thursday, April 7
7:00 am Registration and Morning Coffee
10:00 Coffee Break in the Exhibit Hall and Poster Competition
10:30 Chairperson’s Opening Remarks
Sastry Chilukuri, Partner, Pharmaceuticals & Medical Products, McKinsey & Company
10:40 Transforming Use of Real World Data Analytics
Minnie Chou, Director, Information Systems,
Real world data (RWD) analytics is a key enabler for bringing
effective and safe medicines to patients faster and cheaper. It can improve
study designs, reduce trail enrollment times, facilitate fast-track filing
strategies, shorten response time to health authority queries, and support
value proposition of medicines. This presentation will discuss our approach and
learnings unlocking the power of real world data assets.
11:10 Virtual Systems Pharmacology – The Next Generation of a
TR&D Modeling and Simulation Environment
Marko Miladinov, Informatics Lead, Bristol-Myers Squibb
The internally developed Virtual Systems Pharmacology (ViSP)
platform was implemented at BMS as a dynamic, highly scalable, model agnostic
and therapeutic area agnostic application. The system seamlessly integrates the
modelling tool of choice by the user, a web-based application, command line
utilities, a database back-end and automatically scaling HPC environment built
in the BMS Research Cloud environment that can be used to configure, manage and
execute large-scale simulations for multiple models (of any sort) by multiple
11:40 The New World: Improving Patient Lives through Clinical Analytics and Real World Evidence
Sastry Chilukuri, Partner, Pharmaceuticals & Medical Products, McKinsey & Company
Jonathan Usuka, Knowledge Expert, Pharmaceuticals & Medical Products, McKinsey & Company
Unprecedented access to RWE is unlocking new insights into treatment & precision medicine, with implications for shifting value in the competitive pharma development ecosystem. A detailed understanding of how a patient will respond to therapy is complex & requires significant clinical trials investment, often leading to failure & frustration. But hidden in medical claims data are clues to predict & demonstrate therapeutic benefit. How will it be used to create value & possibly to replace pharma R&D? We examine just that.
11:55 Global Specimen Solutions - Technology Facilitating Unlocking the Future of Medicine™
Peter Tearle, Head of IT Architecture, Global Specimen Solutions, Inc.
12:10 pm Session Break
12:20 Luncheon Presentation I: Why is Data Integration so Hard?
Tim Miller, VP, Integrated Applications, Thomson Reuters
Innovation is the life blood of the pharmaceutical industry and innovation runs on data. The Life Sciences industry is blessed with some of the best data resources – commercial and public databases, literature and ontologies. The Big Data revolution has given us a plethora of information and astounding tools for working with data, yet, the most common single issue faced by researchers is the problem of data integration.
12:50pm Luncheon Presentation II (Sponsorship Opportunity Available) or Lunch on Your Own
1:20 Dessert Refreshment Break in the Exhibit Hall with Poster
1:55 Chairperson’s Remarks
Farida Kopti, Ph.D., Director, Chemistry, Pharmacology, HTS
Informatics & IT, Merck & Co., Inc.
2:00 Harnessing Edge Informatics to Accelerate Collaboration
Tom Plasterer, Ph.D., US Cross-Science Director, Research
& Development Information, AstraZeneca
As scientists in the life sciences we are trained to pursue singular goals around a publication or a validated target or a drug submission. Our failure rates are exceedingly high especially as we move closer to patients in the attempt to collect sufficient clinical evidence to demonstrate the value of novel therapeutics. This wastes resources as well as time for patients depending upon us for the next breakthrough. Edge Informatics is an approach to ameliorate these failures. Using both technical and social solutions together knowledge can be shared and leveraged across the drug development process. This is accomplished by making data assets discoverable, accessible, self-described, reusable and annotatable. The Open PHACTS project pioneered this approach and has provided a number of the technical and social solutions to enable Edge Informatics. A number of pre-competitive consortia and some content providers have also embraced this approach, facilitating networks of collaborators within and outside a given organization. When taken together more accurate, timely and inclusive decision-making is fostered.
2:30 A Platform Strategy for Research
Farida Kopti, Ph.D., Director, Chemistry/Pharmacology/HTS
Informatics & IT, Merck & Co., Inc.
Merck is designing an open, cloud-based, integrated research
data capture, management, and analytics platform to drive operational
efficiency, improved user experience, scientific collaboration (internal and
external) and accelerated decision-making. The objective is to enhance
reusability of data and scientific informatics capabilities by standardizing
data capture and management, and creating an application ecosystem that enables
rapidly advancing science, while reducing the total cost of operations.
3:00 Integrative, Automated Assessment of Human Genetic
Evidence to Enable Decision Making for Drug Target Identification and
Janna Hutz, Ph.D., Director, Head of Human Biology & Data Science, Andover Product Creation Innovation System, Eisai, Inc.
Emerging data have demonstrated the utility of leveraging human genetic evidence to select drug target-indication pairings with increased odds of success in the clinic. Despite this, accessing and interpreting human genetic associations with complex traits is often a manual, labor intensive process. Eisai has established bioinformatic systems for capturing and integrating human genetic data from a variety of study types (GWAS, sequencing, familial, candidate gene, etc.). Automated integration of genetic associations with functional data enables the delivery of summary reports and calibrated numeric scores to biologists and geneticists alike, who are using this information to drive portfolio-level decision-making on selection of targets, biomarkers, and indications.
3:30 Managing Controlled Substances and Other Liability Flags
Roman Affentranger, Head, Small Molecule Discovery Workflows,
Following up on the pre-work done by the Pistoia Alliance, we
implemented a comprehensive solution managing our extensive set of liability
flags for compounds including e.g. narcotic substances.
4:00 Conference Adjourns