July 24-26, 2018

Boston, USA

Day One
Wednesday, 25th July, 2018

Day Two
Thursday, 26th July, 2018

Registration & Breakfast

Chair’s Opening Remarks

Robust Translation from Discovery to Clinic & Early Clinical Design


Introduction and Purpose:

The ultimate goal of researchers is to translate scientific findings into practical clinical applications. Despite successful pre-clinical testing, the majority of early clinical trials for novel drugs fail and of those that survive through to phase III only half become approved for clinical use. AI & machine learning have the potential to assist researchers and clinicians to achieve successful translation and ultimately reduce clinical trial failure rates.


Case Study: GANs’n’Roses: GANs Generating De Novo Molecular Structures with Experimental Validation


  • The deep-neural networks (DNN) are being successfully applied for de novo design of drug-like lead compounds that have the higher probability of success in clinical trials due to the prediction of appropriate physicochemical properties for specific targets


  • The Adversarial Threshold Neural Computer (ATNC) based on generative adversarial network (GAN) architecture and reinforcement learning (RL) has shown the high efficiency in the generating of drug-likeness properties for novel small organic molecules


  • In vitro validation of the molecules generated by ATNC confirmed that ATNC is an effective method for producing hit compounds

Case Study: Reduction of High-dimensional Clinical RNA-seq Data by Deep Learning: Methods and Its Practical Application

  • Shanrong Zhao Director, Computational Biology- Worldwide Research & Development, Pfizer


  • Challenges of high-dimensional transcriptomic data in machine learning


  • Methods for reduction of high-dimensional transcriptomic data


  • Deep learning based approaches for reduction of clinical RNA-seq data


Case Study: Using Machine Learning to Analyze Clinical Trials that Fail to Meet Primary Endpoints


  • Factors that cause clinical trials to fail their primary endpoints can be difficult to uncover with traditional statistical methods. Modern Machine Learning techniques can discover features which cause the primary endpoints to fail in historical clinical trial data


  • Machine Learning methods can scale longitudinally across studies and seamlessly handle a large number of covariates


  • Insights on features that drive primary endpoint failure can be used to inform better clinical trial study design in the future

Case Study: Applying AI-Driven Approaches to Improve Clinical Development Efficiency


  • Reviewing the need to implement AI in Clinical Development


  • Discussing the key points necessary to deploy a functional AI-guided clinical trial


  • Demonstrating how AI can de-risk clinical development by stratifying patients based on selected outcomes


Speed Networking & Morning Refreshments

Matching AI & Machine Learning Benefits to Pharma Challenges in Clinical Development


Introduction and Purpose:

Failed clinical trials result in wasted money and resources, lost jobs and a setback in research. The financial pain of trial failure will eventually encourage pharma to change its approach and processes to be more efficient and successful in the long term. While the use of AI in clinical development is still limited, identifying, addressing and overcoming the pitfalls and opportunities associated with its real world applications are the enabler to the success of AI in drug development.


Case Study: The Future is Now– The Most Promising Areas of AI/Machine Learning for Pharmaceutical Industry

  • Haoda Fu Research Advisor & Enterprise Lead for Machine Learning and Artificial Intelligence Group, Eli Lilly


  • There are growing interests on AI for clinical development, pitfalls and opportunities co-exist. Where are the real opportunities?


  • Through years and extensive evaluations, we have identified 4 areas which are tangible and have potential significant impacts on drug development


  • Successful examples will be provided and potential next steps will be discussed


Case Study: Pharma’s Holy Grail: How to Solve the Right Patient Stratification Problem?


  • Improving the success of traditional drug development


  • The right types of datasets


  • Lantern Pharma Approach: A Specific AI Application Technology


Case Study: Real talk: 5 Ways to Succeed with Big Pharma


  • What are the common themes of successful AI partners? What are the most common mistakes?


  • What are the unique challenges of partnering with big pharma?


  • What is Sanofi’s approach to digital partnerships?


Moving from Theory into Practice by Implementation of AI in clinical Development

Networking Lunch

Case Study: From Science to Practice: Using Machine Learning in Production to Optimize Clinical Trials


  • Challenges in creating representative training datasets for building robust machine learning (ML) models
  • Leveraging ML models to answer questions in feasibility, study startup, patient recruitment & retention, risk monitoring and compliance
  • Verification and validation of machine learning based solutions in controlled regulatory environments and monitoring ongoing performance

Discuss: Real World/Pragmatic Clinical Trials and the Use of AI to Design & Monitor Them:

  • Shanrong Zhao Director, Computational Biology- Worldwide Research & Development, Pfizer
  • Ray Liu Senior Director & Head, Advanced Analytics & Statistical Consultation, Takeda
  • Raj Bandar Head of Data Sciences & Analytics , Bill & Melinda Gates Medical Research Institute
  • Sean Grullon Machine Learning Data Scientist, GSK


  • How to leverage AI as an informatics tool to inform whether drugs work as well in the real world as they did in clinical trials?


  • How AI could help with choosing patients that are true representation of the population?


  • What are the limitations of monitoring an AI-driven pragmatic clinical trial and how to over come them?


Case Study: How to Maximise the Benefits of Integrating AI in Clinical Trials to Improve Recruitment Efficiency


  • Better adherence = better clinical trials


  • Connecting the dots faster, better & smarter

Afternoon Refreshments & Networking

Roundtable Discussions: Our breakout roundtables will allow you to have more intimate discussions with AI and pharma leaders around some of the hottest topics in the field. Discover multiple perspectives on these key issues, so that you can learn from your fellow experts in the audience. Drive your own learning, crowd-source ideas and get inspired. Immerse yourself in the following discussions:


Roundtable Topics:


1.       Improve patient enrollment through predictive modeling


2.       Using machine learning to improve trial efficiency


Chair’s Closing Remarks & End of Day One