February 26-28, 2018

San Francisco, USA

Day One
Tuesday 27th February, 2018

Day Two
Wednesday 28th February, 2018

Breakfast & Networking

Chair’s Opening Remarks

Enhancing Drug Discovery Productivity Using AI & Machine Learning Technologies

Case Study: Massive Scale Machine Learning for Integrative Compound Activity Prediction

  • Hugo Ceulemans Scientific Director Discovery-Data Sciences, Johnson & Johnson


  • Multitask learning at industrial scale boosts predictive performance
  • Enlisting alternative data sources like cellular images and transcriptional profiles for compound boosts predictive performance
  • Towards collaborative learning from private data

Case Study: Pure AI Approach Versus Augmented Intelligence Approach: Augmentation of Data Using AI to Predict Missing Data to Consequently Build Disease Models


  • How to optimize the decision making process?
  • Is cognitive technology they way forward?
  • How to prepare for augmented intelligence as an emerging technology?

Case Study:The In Silico Biosciences Quantitative Systems Pharmacology (QSP) Platform as the Next Step in the Big Data AI Value Chain in CNS


  • Bridging the gap between big data AI Targets and clinical phenotypes with QSP Augmented intelligence technology
  • Optimizing in silico biomarkers as a quantitative target discovery and target validation engine, especially for multi-target drugs
  • Using the platform downstream in compound and clinical development

Case Study: Discovery of New Chemical Entities Using Network-driven Drug Discovery Augmented by AI

  • Ben Allen Senior Computational Research Scientist, e-Therapeutics


  • What is network driven drug discovery?
  • How does AI/machine learning contribute to this approach?
  • Validation case studies

Practical Applications of AI & Machine Learning in Drug Discovery

Case Study: Accutar Biotech Employs Artificial Intelligence to Revolutionize Drug Discovery

  • Jie Fan CEO, Accutar Biotechnology


What is our philosophy?

  • To derive a data-driven principle that has the power of explaining physical and
    chemical nature of biological systems, which we harness to accelerate drug discoveries
    What have we done?
  • A data-driven atom-based scoring function is learned from 100,000 protein crystal
    structures containing information of >100 million amino acid side chains
  • A dynamic deep neural network specifically designed for chemical informatics

What have we done?

  • A data-driven atom-based scoring function is learned from 100,000 protein crystal structures containing information of >100 million amino acid side chains
  • A dynamic deep neural network specifically designed for chemical informatics

Why is our work impactful?

  • Beating current gold standard in computation-aided drug discovery
  • Drug pocket side chain conformation prediction and drug docking with significantly
    increased accuracy compared to standard method (Schrodinger)
  • Prediction of chemical compound characteristics (solubility, binding affinity, etc)
    significantly better than current standard

Morning Refreshments & Networking

Case Study: Harnessing the Power of AI to Improve Translation: From Biology to Discovery and Discovery to Development


  • Novel molecule modulation for identified targets
  • Optimizing molecule properties/profiles to enhance productivity
  • PK/PD modelling & simulation
  • Evaluating compound’s ADMET characteristics

Case Study: Multi-dimensional Drug Discovery with AI-enabled Phenomics

  • Ron Alfa VP, Discovery & Product, Recursion Pharmaceuticals


  • Combining AI methods with high throughput phenomics enables drug discovery at a radical pace
  • Cellular images contains a wealth of biological information that can be used for drug discovery campaigns and hit prioritization
  • Recursion is systematically mapping human biology through images

Current Focus of AI & Machine Learning Technologies in Drug Discovery

Case Study: The Role of Structural Analysis in the Improvement of AI Methods Used in Drug Design


  • Current status and challenges for the application of AI in drug discovery
  • Describing the methods based on structural analysis for the generation of protein fingerprints and how that contributes to small molecule drug discovery
  • Challenges for the application of AI methods in the development of therapeutic antibodies

Networking Lunch

Case Study: Accurate Data and Domain Expertise: Key Ingredients of AI-driven Target and Drug Discovery

  • Tudor Oprea Professor-Medicine & Chief of Translational Informatics & Internal Medicine, University of New Mexico


  • Illuminating the Druggable Genome is an NIH project focused on accurate data wrangling, processing and analytics for target discovery
  • DrugCentral, an on-line compendium for drugs and drug targets, can serve as basis for understanding the molecular basis of therapeutic action
  • Drug-, protein- and disease- knowledge graphs can be used to enable AI systems in drug discovery

What the Future of AI & Machine Learning in Drug Discovery Holds: The Emerging Trends

Case Study: Machine Learning for Safety Risk Models in Development

  • Gregory Bell VP, Global Therapeutic Area Head, Immunology, Infectious Disease and Ophthalmology, PD Safety Science, Genentech


  • Leveraging what we know about the drug & the available data
  • The use of pharmacovigilance data in AI-driven drug repurposing approaches
  • Developing risk models with or without machine learning: Almost no one does this explicitly

Afternoon Refreshments & Networking

Case Study: The Application of AI & Machine Learning in Multi-target Drug Discovery: Polypharmacology


  • In silico approaches for polypharmacology and its application
  • Challenges of integrating big data in biology
  • Application of knowledge and algorithms for drug discovery

Case Study: Deep Learning in Pharmacogenomics: From Drug Discovery to Patient Stratification

  • Gerald A. Higgins Research Professor, Computational Medicine and Bioinformatics, University of Michigan Medical School


  • Promising applications of pharmacogenomics drug discovery and development as well as medication optimization
  • Identification of regulatory pharmacogenomic variants and drug target discovery using deep learning
  • The predictive power of machine learning is realized when it is combined with prior domain knowledge, such as gene networks and pharmacodynamic pathways

Chair’s Closing Remarks

Close of Day Two & End of AI Pharma Innovation: Drug Discovery Summit 2018