AI in Pharma

Investment and the application of AI in the pharmaceutical industry is rapidly gaining momentum. We bring together CEOs, CIOs, CTOs and Global AI, IT and Informatics Experts from leading pharmaceutical and technology companies to give strategic and visionary talks based on use cases throughout the drug development pipeline.

Tuesday, December 4, 2018

7:45 am Registration Open

8:20 am - 12:00 pm AI World Plenary Session - View Details

11:00 am - 2:00 pm Concession Stand Open for Lunch in the Exhibit Hall

Track Program

Henstock_PeterTrack Chair

Peter Henstock, PhD, AI & Machine Learning Lead, Pfizer

Protabit 1:15 ProtaBank: AI and Machine Learning for Protein Engineering. Where's the Data?

Olafson_BarryBarry Olafson, PhD, CEO, Protabit LLC

A significant challenge for applying AI and machine learning to protein engineering is accumulating the needed protein sequence mutation data over multiple properties in a central repository. The scientific community has created the GenBank database for nucleic acid sequences and the PDB database for protein structures and now is the time to expand the protein sequence mutation database, ProtaBank. We will describe our current efforts and suggest how industry can contribute through a pre-competitive collaboration.

1:40 The Potential for Advanced Analytics in the Pharmaceutical Industry

Ringel_MichaelMichael Ringel, PhD, Senior Partner and Managing Director, Boston Consulting Group (BCG)

Advanced analytics, including artificial intelligence, have the potential to improve any decision. In pharma, biopharma and healthcare more broadly, this offers the promise of a major improvement in cost and effectiveness, given that an enormous portion of current spend in both domains ends up as waste. In this overview presentation, we set up the opportunities, challenges and reasons for optimism.

2:00 CO-PRESENTATION: Artificial Intelligence Accelerants in Oncology Informatics

Data and information flows in oncology research provide an ideal opportunity to evaluate emerging applications in Artificial Intelligence.  While some bioinformatics and next-generation sequencing workflows are becoming standardized and can provide a well-described input dataset for AI applications, other require extensive data preparation and stewardship prior to exploitation.  Applications for both well-structured, well-stewarded and poorly-structured data will be described.

Plasterer_TomTom Plasterer, PhD, US Cross-Science Director, R&D Information, AstraZeneca

Dry_JonathanJonathan Dry, PhD, Director of Bioinformatics, Oncology, AstraZeneca

2:20 Why Is the Pharmaceutical Industry Finding AI so Difficult?

Henstock_PeterPeter Henstock, AI & Machine Learning Lead, Pfizer

The rise of AI has dramatically changed the course of many industries as they have been able to adopt this range of new technologies. Although pharmaceutical companies regularly incorporate technology innovations, the transformation has arguably been much slower. This presentation will outline some of the challenges, highlight some successful examples, and describe some of the strategies we are using to drive AI across the organization.

2:40 Data Store: Making all Data Machine Learnable at NIBR

Walker_JohnJohn R. Walker, Head of Core Data and Analytics, NIBR Informatics, Novartis 

NIBR invests billions each year in generating scientific data. The Data Store is the central platform to find, store and process these data assets. This presentation will describe the challenges, strategy and implementation details for this new platform at Novartis.


3:00 Refreshment Break in the Exhibit Hall with Poster Viewing

3:45 FEATURED PRESENTATION: Integrating AI Workflows in Drug Discovery

Baldoni_JohnJohn Baldoni, PhD, Senior Vice President, In silico Drug Discovery, GlaxoSmithKline

GSK formed an In-Silico Drug Discovery Unit to explore how AI/ML/DL methods can affect the drug discovery process in multiple dimensions.  The unit takes advantage of existing internal capability and integrates subject area expertise from external pharma-focused companies and non-pharma sectors to improve the effectiveness of drug discovery. The goal of the unit is to flexibly use and integrate these platform capabilities in its work to discovery a drug. The talk will describe the enabling activities that have emerged outside of the company and those within the company to enable this approach, the business model that is forming and successes and failures to date.

4:05 How Models in Drug Discovery can Leverage Deep-Learning: Predictions and More

Neumann_EricEric Neumann, PhD, CEO & Founder, Aidaka

4:25 PANEL: The Use of AI to Disrupt Drug Discovery: How to Reduce Time and Costs and Increase Throughput

  • Where can Deep Learning show benefits in R&D?
  • Algorithms: Data distribution vs. outcome prediction
  • Critical need in DL for transparency and solution interpretability
  • Deep learning works great on image data.  What about molecular data which is sparse, dirty, inconsistent?  Drug Discovery must be accurate on outliers and deep learning is not so great here.  What do we do?
  • Application of machine learning to toxicity and Pk
  • Application of machine learning to reduce computational time of computational chemistry
  • Augmented intelligence
  • Target discovery and lead discovery require different methods
  • How can ML/AI approaches be used to speed up the drug discovery process: Target ID, lead optimization and clinical development
  • Do we have access to the right data, suitable for state of the art ML/AI approaches?
  • How can we recruit the best given the high level of competition across multiple sectors? 

Zhavoronkov_AlexModerator: Alex Zhavoronkov, PhD, Founder, Insilico Medicine


Davies_MarkMark Davies, Vice President Biomedical Informatics, Benevolent AI

Brase_JimJim Brase, Chief Technical Officer & Interim Co-Lead, ATOM and Deputy Associate Director for Computation, Lawrence Livermore National Laboratory

Akmaev_SlavaSlava Akmaev, PhD, Senior Vice President & Chief Analytics Officer at BER

Neumann_EricEric Neumann, PhD, CEO & Founder, Aidaka

Addison_EdEd Addison, CEO, Cloud Pharmaceuticals

5:10 - 5:30 Plenary Keynote Presentation - View Details

5:30 - 6:30 Networking Reception in the Exhibit Hall with Poster Viewing


Wednesday, December 5

7:45 am Registration Open

8:20 am - 12:45 pm AI World Plenary Session -View Details

11:00 am - 2:00 pm Concession Stand Open for Lunch in the Exhibit Hall

Track Program


Dany De Grave, Senior Director, Innovation Programs and External Networks, Sanofi

2:00 Artificial Intelligence for the Clinical Study Report

Gidh-Jain_MadhaviMadhavi Gidh-Jain, PhD, Senior Director, Head Medical Writing, Clinical Documentation, Sanofi

The goal of this global Sanofi partnership with YESOP (2016 Technology Innovation Awards Winner) is to build an AI engine to create well written, meaningful narrative for clinical study reports. This will gain efficiency by reducing time to first draft while decreasing quality control (QC) and review requirements. Generating text that is standard, for example, safety sections of the CSR will allow the medical writer to focus on the more complex interpretation of data while reducing the time required to draft. There is also efficiency gained by reducing review and quality control (QC) time. Our proof of concept for AI - generating the section of the CSR describing the brief summary of adverse events, has provided us with impetus to proceed with the next phase of the pilot. Currently, we are in the process of completing the AI – generation of the safety section. Conceptually, the whole value of AI is writing an algorithm that accurately predicts and smart searches through data. The AI system is source code, and its actions will only ever follow from the execution of the instructions that we initiate. Therefore, the ability of the AI to make high-quality documents rests on the learning medical writers provide. Our project, AI4CSR will provide an opportunity for faster CSR writing, accelerating time to submission, ultimately resulting in patients getting vital medicine more quickly.

2:25 Strategy and Application of AI in Closing the Loop from Clinical Trials to Discovery Biology

Cho_CarolynCarolyn Cho, PhD, Director, Immunology Therapeutic Area Pharmacometrics Lead, Merck

Systems immunology is a powerful approach to understanding complex pathophysiology in immunological disorders and in vaccine response. However, the identification of testable causal hypotheses has been widely recognized as one of the most significant challenges of this approach. Causal modelling of data from a previous “bioage” clinical study (Fourati et al., Nat Comm 7:10369, 2016) is being applied to identify hypothesized mechanisms of vaccine hyporesponse in the elderly.

2:45 Generating a Business Case for Innovation: How Can AI Enable the Future of Clinical Trials?

Gummadi BaskerBasker Gummadi, Digital Innovation Lead, Bayer

How can driving innovation enable and achieve business objectives? This talk will highlight business cases within clinical trials that can be achieved by AI, Robotic Process Automation and Blockchain.


3:05 Refreshment Break in the Exhibit Hall - Last Chance for Viewing

3:40 PANEL: Can AI & ML Make Clinical Trials Faster, More Effective and Patient Centric?

  • AI: The tangible value beyond the hype
  • Applying AI frameworks to improve trial outcomes
  • Leveraging data for machine learning projects
  • Keeping the focus on the patients
  • Knowledge engineering (AI, ML, and DL) in pharma R&D
  • New ethical considerations for the life sciences industry
  • What is the future for consent and permissions? How granular should this be when patients control their data?
  • How we feed the machine is as important as the machine itself
  • How is the role of the health tech ethicist different? How can companies build that capacity?

Holoubek_SaraSara Holoubek, CEO, Luminary Labs


Slater_TedTed Slater, Global Head, Scientific AI & Analytics, Cray

Flink_BalazsBalazs Flink, MD, Clinical Trial Analytics Lead, R&D Business Insights and Analytics, Bristol-Myers Squibb

Bouton_ChristopherChris Bouton, PhD, CEO, Vyasa

4:20 Exploring New Ways of Working in Regulatory: Sanofi's Proof of Concept Pilot

DE_GRAVE_DANYDany De Grave, Senior Director, Innovation Programs and External Networks, Sanofi

At Sanofi we defined AI for the Regulatory Space as “The ability of an intelligent system to quickly integrate vast amounts of diverse, complex and unstructured regulatory-pertinent information transforming it into actionable insights for strategy planning and decision making throughout the product lifecycle. It allows iterations at an unprecedented scale in human-like terms and provides “logic” for insights provided.” We will share lessons learned from a Proof of Concept pilot.

4:40 The Use of AI in Real World Evidence at Sanofi 

Cliona Molony, PhD, Head of Advanced Analytics, Real World Evidence & Clinical Outcomes, Sanofi

5:00 Close of AI World 2018