Original Agenda
We are actively working with our speakers to confirm their availability for our new dates. Initial response from our speakers has been very positive, and we are optimistic we will have the new programs ready to share here soon.

Artificial Intelligence for Early Drug Discovery, August 26-27 2020, San Diego, CA

Cambridge Healthtech Institute’s 2nd Annual

Artificial Intelligence for Early Drug Discovery

AI & Machine Learning for Drug Design and Lead Optimization

NEW DATES - AUGUST 26-27, 2020

The Artificial Intelligence for Early Drug Discovery conference will bring together a diverse group of experts from bioinformatics, chemistry, target discovery, DMPK, and toxicology to talk about the increasing use of computational tools, artificial intelligence (AI) models, machine learning (ML) algorithms, and data mining in preclinical drug development. Starting with an overview of current challenges and opportunities, the talks will highlight how AI/ML can help with drug design, target identification, lead optimization, PK/PD predictions, and early safety assessments. The speakers will offer insights into the caveats and limitations of AI/ML-based decision-making using relevant case studies and research findings. The conference will offer an excellent opportunity to network and share ideas and best practices.

Final Agenda


Wednesday, aPRIL 15

12:30 pm Registration Open

12:45 Dessert Break in the Exhibit Hall with Poster Viewing (Sponsorship Opportunity Available)


1:30 Welcome Remarks

Tanuja Koppal, PhD, Senior Conference Director, Cambridge Healthtech Institute

1:35 Chairperson’s Opening Remarks

Yuan Wang, PhD, Senior Principal Scientist, Head of TPD Data Science, UCB Pharma

1:40 Applied Machine Learning in Compound Mechanism Deconvolution

Yuan WangYuan Wang, PhD, Senior Principal Scientist, Head of TPD Data Science, UCB Pharma

Modern drug discovery calls for increased use of phenotypic screens and novel targets and modalities are being explored in the process. The use of potent and selective chemical tools (probes) in phenotypic screens can help understand underlying biological processes or help deconvolute unknown mechanisms. We have used computational analytics and machine learning models to: 1) select tool compounds; 2) predict targets; and 3) design better sets of compounds.

2:10 AI Powered Design of Small Molecules Accelerated by Structure-based Constraints

Bradley_AnthonyAnthony Bradley, PhD, Head of Structural Bioinformatics, Exscientia

Structural data are central to many of our projects since they enable clear and directed hypothesis and constraint generation. In this talk we present three ways in which they empower Centaur ChemistTM through examples from our fully validated design engine. First, we show how our best-in-class AI systems use 3D ligand and protein-ligand structures to accelerate compound design. Second, we present how feedback between 3D and 2D models enable identification of gaps in data. Finally, we outline how a new breed of Deep Neural Networks and Generative Models promise a further step-change in productivity.

2:40 Deep Generative Autopilot for the Real-World Design of Novel Lead Compounds

Sang Ok Song, PhD, Co-Founder and Chief Transformation Officer, Standigm, Inc.

Standigm has applied deep generative models to design novel therapeutic compounds and launched Standigm BEST®, a proprietary molecular generative platform for lead discovery and optimization. On top of the main molecular generative algorithm, we developed an automated molecular design workflow to optimize and prioritize machine generated compounds for further synthesis and experimental validation. The most recent progress including real-life case studies will be shared.

3:10 Structure-guided De Novo Drug Design Via AI/deep Generative Modeling

ann Gaston-MathYann Gaston-Mathé, Co-Founder, CEO, Iktos

Artificial Intelligence is an emerging research field in the context of drug discovery. Complementing its ligand-based deep generative modeling technology for de novo drug design, Iktos has developed structure-guided deep generative modeling technology that overcomes the applicability domain limitations of ligand-based methods. We will present this novel technology and a case study where the structure-guided generation of small molecules enabled the discovery of AI-designed molecules similar to known ligands.

3:40 Refreshment Break and Book Signing in the Exhibit Hall with Poster Viewing (Sponsorship Opportunity Available)


4:30 FEATURED PRESENTATION: Benefits, Limitations and Diversity of AI Models in Drug and Target Discovery

Ruben Abagyan, PhD, Professor, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego

Computer models that are capable of predicting several thousands of biological activities for any chemical along with their ADMET properties have improved dramatically with the rapid growth of experimental data. The resulting network, illustrated by cancer drugs, has an extensive multi-target profile for each drug. These models use different mathematical methods, and help to predict new targets for known compounds, repurpose to new indications, search for compounds with specific multi-target profile, or identify potential liabilities.

5:30 Breakout Discussions - View All Breakouts

In this session, attendees choose a specific roundtable discussion to join. Each group has a moderator to ensure focused conversations around key issues within the topic. The small group format allows participants to informally meet potential collaborators, share examples from their work, and discuss ideas with peers.

Topic: AI-Driven Target Discovery and Therapies

Moderator: Ruben Abagyan, PhD, Professor, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California San Diego

  • Types of AI models predicting individual target activities of small molecules
  • May the docking be a useful intermediate step before the AI model is applied?
  • How under-characterized is the set of activities of small molecule therapeutics and drug candidates?

Topic: Use of Modeling Tools and Strategies for Predicting ADME-Tox Properties

Moderator: Maria A. Miteva, PhD, Research Director at Inserm, Medicinal Chemistry and Translational Research

  • Machine-learning and structure-based approaches for ADME-Tox prediction
  • Deep learning for ADME-Tox prediction
  • Precision medicine and ADME-Tox

Topic: Trends in AI for Accelerating Drug Discovery

Moderator: Amol Jadhav, PhD, Industry Consultant, Transformational Health, Frost & Sullivan

  • Current trends for the application of AI towards preclinical drug discovery, status and challenges
  • What measures should be taken to invest and apply AI at various stages of drug development?
  • Industry-Academia partnerships, shared experience from startups, academia and impact assessment

Topic: Binding Datasets in AI for Drug Discovery

Moderator: Kaspar Cottier, PhD, CTO, Creoptix

  • How can actual measurement data from binding studies contribute to the success of AI for drug discovery?
  • How does data protection within the industry affects the development of AI for drug discovery? What role can academic institutions take for filling in data gaps?
  • Which parameters (kinetics, affinity, thermodynamics…) are most important, and how can instrument manufacturers contribute to future developments?

6:15 Close of Day


6:30 Dinner Short Courses

Thursday, April 16

8:00 am Breakfast Plenary Technology Spotlight (Sponsorship Opportunity Available) or Morning Coffee

8:45 Plenary Welcome Remarks from Event Director with Poster Finalists Announced

Anjani Shah, PhD, Senior Conference Director, Cambridge Healthtech Institute

8:55 Plenary Keynote Introduction (Sponsorship Opportunity Available)



baranPphilTranslational Chemistry

Phil Baran, PhD, Professor, Department of Chemistry, Scripps Research

There can be no more noble undertaking than the invention of medicines. Chemists that make up the engine of drug discovery are facing incredible pressure to do more with less in a highly restrictive and regulated process that is destined for failure more than 95% of the time. How can academic chemists working on natural products help these heroes of drug discovery – those in the pharmaceutical industry? With selected examples from our lab and others, this talk will focus on that question highlighting interesting findings in fundamental chemistry and new approaches to scalable chemical synthesis.

9:45 Coffee Break in the Exhibit Hall with Poster Viewing (Sponsorship Opportunity Available)


10:40 Chairperson’s Remarks

Maria A. Miteva, PhD, Research Director, Molécules Thérapeutiques in silico (MTi), Inserm Institute

10:45 Integration of Structure-Based and Machine Learning Approaches to Predict Inhibition of Drug Metabolizing Enzymes

Maria MitevaMaria A. Miteva, PhD, Research Director, Molécules Thérapeutiques in silico (MTi), Inserm Institute

We will present in silico approach to predict inhibition of drug metabolizing enzymes. We focus on Cytochrome P450 (CYP) responsible for the metabolism of 90% drugs and on sulfotransferase (SULT), a conjugate Phase II metabolizing enzyme. We developed an original in silico approach for the prediction of CYP2C9 and SULT1A1 inhibition combining protein structure knowledge, dynamic behaviors in response to ligand binding and modern machine learning modeling.

11:15 Leveraging Image-Derived Phenotypic Measurements for Drug-Target Interaction Predictions

Arvind RaoArvind Rao, PhD, Associate Professor, Department of Computational Medicine and Bioinformatics, University of Michigan

We propose a novel in silico drug discovery approach to identify kinase targets that impinge on nuclear receptor signaling with data generated using high-content analysis (HCA). Using imaging-derived descriptors, we provide prediction results of drug-kinase-target interactions based on single-task learning, multi-task learning, and collaborative filtering methods. These results suggest that imaging-based information can be used as an additional source of information for existing virtual screening methods, thereby making drug discovery more efficient.

11:45 Complexity Simplified - Comprehensive Data Management for Complex World

Heather Mattson Arnaiz, PhD, Customer Engagement Scientist, Collaborative Drug Discovery

CDD brings drug discovery full circle. From unstructured experiments to structured registrations, we ensure data is accessible from a single, secure location.  CDD Vault fits with your informatics and AI tools as the platform for managing data.  This is easily accomplished within CDD Vault that includes an API for automation. 

12:00 pm Explainable Biology for Improved Therapies

Igor JurisicaIgor Jurisica, PhD, DrSc, Senior Scientist, Krembil Research Institute; Professor, University of Toronto; Visiting Scientist, IBM CAS

Integrative computational biology and AI help improve treatment of complex diseases by building explainable models. From systematic data analysis to improved biomarkers, drug mechanism of action, and patient selection, such analyses influence multiple steps of drug discovery pipeline. Data mining, machine learning, graph theory and advanced visualization help with accurate predictions, making disease modeling more comprehensive by intertwining computational prediction and modeling with biological experiments.

12:30 Session Break

12:40 Luncheon Presentation (Sponsorship Opportunity Available) or Enjoy Lunch on Your Own

1:30 Dessert Break in the Exhibit Hall with Poster Awards Announced (Sponsorship Opportunity Available)


2:15 Chairperson’s Remarks

Wenjin Zhou, PhD, Assistant Professor, Department of Computer Science, University of Massachusetts Lowell

2:20 New Anti-Cancer Peptide Design Using a GAN-Based Deep Learning Method

Wenjin Zhou, PhD, Assistant Professor, Department of Computer Science, University of Massachusetts Lowell

Cancer is a deadly disease that causes an estimated 9.6 million deaths a year. Pharmaceutical drugs are important, but developing new drugs is difficult and expensive. Here we generate a new peptide for PD-1, which is closely linked to a wide variety of cancers, using a new application called GANDALF to design new peptides. We present a peptide generated by our prototype to bind with PD1 and compare it to FDA approved drugs and results from a comparable method, Pepcomposer.

2:50 AI for Accelerated Pre-Clinical Drug Discovery: From Data Mining to Screening Automation

Amol_JadhavAmol Jadhav, PhD, Industry Consultant, Transformational Health, Frost & Sullivan

This presentation will focus around the role of life sciences big data, technologies and emerging application of AI for the early drug discovery. The case studies discussing the impact of automation and miniaturization approaches coupled with machine learning on the speed and efficiency for the compound screening will be discussed. Additionally, a brief assessment of the therapeutics area-wise activity, current trends will be provided.

3:20 FEATURED PRESENTATION: Artificial Intelligence and Small-Molecule Drug Metabolism

S. Joshua Swamidass, MD, PhD, Assistant Professor, Immunology and Pathology, Laboratory and Genomic Medicine; Faculty Lead, Translational Informatics, Institute for Informatics, Washington University

We have been building artificial intelligence models of metabolism and reactivity. Metabolism can both render toxic molecules safe and safe molecules toxic. The artificial intelligence models we use quantitatively summarize the knowledge from thousands of published studies. The hope is that we could more accurately model the properties of medicines, to determine whether metabolism renders drugs toxic or safe. This is just one of many places where artificial intelligence could give traction on the difficult questions facing the industry.

3:50 Networking Refreshment Break

LabTwin 4:20 How Voice Assistants are Empowering Drug-Discovery Scientists?

Guru SinghGuru Singh, Head, Growth, LabTwin GmbH

Drug Discovery scientists spend 50-80% of their time at the bench working on experimental protocols. Currently, there is no easy way to record data or access information while performing experiments. Here, new voice-powered digital tools are uniquely positioned to connecting scientists with this capability at the point of experimentation. Application of voice-powered digital assistants holds great promise in R&D, particularly in pharma R&D to improve efficiency by lowering costs, reducing error rates, and improving productivity.

4:50 Applications of Deep Learning-Based Approaches for Non-Target Based Drug Discovery

Keshavarzi_ArashArash Keshavarzi Arshadi, MS, Doctoral Student/Research Fellow, College of Medicine, University of Central Florida

My talk will be about applying DL-based approaches for non-target based virtual screening and drug repurposing. I will describe how DeepMalaria (our patented model) could identify 87.75% of all macrocyclic hits as the first attempt at using AI-based models for malaria drug discovery and repurposing. Also, I will present a model we developed using transfer learning/unsupervised studies for discovery of Anti-Microbial Peptides (AMP) as a state of art approach (AMPDeep).

5:20 Drug Hybridization Approach to Drug Design

Egbe_ClintonClinton Egbe, PharmD, Master’s Student in Drug Design and Discovery, University of Salford

Drug hybridisation approach to drug design is an innovative approach to design drugs that are safe, effective, and affordable. The “Drug Hybridizer” is a bioinformatic tool that will combine in silico molecular fragments of two or more old drugs to produce hybrid compounds that can modify more than one pharmacological target simultaneously for the treatment of multi-pathological disorders such as strokes, cancers and infectious diseases. It will also help to address the problem of drug resistance.

5:40 Close of Conference