AUGMENTED INTELLIGENCE IN DRUG DISCOVERY XCHANGE
September 28, 2023
Welcome to hubXchange’s Augmented Intelligence (AI) in Drug Discovery Xchange West Coast 2023, bringing together executives from pharma and biotech to address and find solutions to the key issues faced in AI-led drug discovery.
Discussion topics will cover Data Quality, Target Identification, Lead Generation & Lead Optimization and Drug Response Prediction.
Take advantage of this unique highly interactive meeting format designed for maximum engagement, collaboration and networking with your peers.
Please note this is an In-Person meeting.
VENUE DETAILS: DoubleTree by Hilton San Francisco Airport, 835 Airport Blvd, Burlingame CA 94010-9949
SNAPSHOTS OF DISCUSSION TOPICS
- Unleashing the power of quality data in AI & drug discovery
- Best practices for building trust worthy data
- Role of AI in target discovery: challenges and opportunities
- Identification of novel targets to drug for a disease of interest
- Expanding and measuring applicability domains in machine learning
- Use of AI/ML in lead optimization (tbc)
- Predicting patient response using machine learning: where are we now?
- Can machine learning methods provide insights and predictions for cancer drug response?
Full Xchange Agenda
Click on each track for detailed agenda
Opening Address & Keynote Presentation by Digital Science
Taming the beast – A one year vision on Generative AI in knowledge discovery on scientific literature and more
Starting in 2019 when the first GPT-2 model entered the stage of Generative AI indicating that something new is happening, followed just a year later by GPT-3 and its powerful model variants and finally with the introduction of ChatGPT and GPT-4 not even a year ago, transformers have completely changed the daily life of any data science and innovation department around the world by today – especially in research and knowledge driven businesses like the life science, chemical and pharmaceutical industries.
This talk will introduce the perspective of Digital Science on this unprecedented time of a fast changing landscape of technology and – by projecting the experiences from the past to the future – will attempt to formulate a vision on where we might stand in one year ahead. Big Tech models are only one aspect of this venture into the near future, as strong movements in the open source community with astonishing developments on transformer models of much more reasonable sizes (still preserving a lot of their language understanding) have opened up alternative and cost efficient approaches for deep analysis and knowledge extraction from large amounts of scientific literature and other text corpora. And in addition symbolic AI, ontologies and the advantages of well structured data are contributing to the trust in and the reliability of our systems when combined with Generative AI.
The Artificial Intelligence “beast” is out and it is extremely powerful with countless opportunities. But “taming” it to effectively augment the human intelligence of domain experts in a trustful, reliable and cost efficient way – not to forget the carbon footprint – is an art in itself and a real challenge for any of us who lives in data science and innovations.
Martin serves as Head of Innovations at Digital Science and manages an agil team of AI engineers and Data Scientists with the mission to translate latest developments in AI, NLP and CL into use cases for the DS portfolio companies. He also serves as a member of the Holtzbrinck Publishing Group committee of AI experts. Prior to Digital Science, Martin was the CTO and Head of Technology at e.Consult AG, a German Legal Tech company to which he also was founder and shareholder and served them as member of the supervisory board for 13 years. Prior to e.Consult AG he served as Chief Technology Officer of Collexis Holdings Inc. SC USA, which was acquired by Reed Elsevier in 2010. While at Collexis, Martin was responsible for the core products and innovations mainly in knowledge management and discovery for the life science and biomedical markets. Prior to Collexis, he was founder, shareholder and managing director of SyynX Solutions GmbH, which was acquired by Collexis Holdings in 2009. With SyynX, Martin invented the largest professional social network for the life sciences community „BiomedExperts“ which reached approx. 500.000 members within one year in 2008. Besides his profession in AI, NLP and innovations he always was and is an enthusiastic musician in avant-garde/free jazz and contemporary classical music and a passionated amateur astronomer and photographer.
Best Practices for Building Trust Worthy Data
- Defining “Data Quality”
- Data Producers vs Data Consumers: who owns the responsibility for Quality Data?
- Incentives and Carrots – successful management approaches to ensuring Data Quality.
- Technologies – successful solutions & approaches to enable validation, metadata, tagging, cleansing/harmonization, etc
Patented start-up scientist, global IT leader, and innovator with hybrid skills in immunology, software development and business architecture. Responsible for Research, Nonclinical Development and Manufacturing Technologies.
Strategy and Execution: Applying successful get-to-market experience to identify right problems and solve with impactful digital solutions. Managing portfolios and products to transform R&D and Early Manufacturing via steering, disciplined portfolio management, POCs, and an agile/waterfall fusion.
Navigating Complexity: Developed 7 enterprise systems across 5 global sites in 3 years through questioning, prioritization, and problem solving. Value delivered for bioprocess machine learning, antibody sequence analytics, biologics registration, system-of-systems design, business workflow automation, and COTS config.
- Maintaining clean data before and after application of AI.
- Importance of FAIR data and standardized assay annotation.
- Where to find quality data to facilitate drug development.
Spotlight Presentation by GERO
Disrupting Disease Landscapes: AI-Driven Insights from Large Biomedical Datasets
This speech will delve into the profound impact of artificial intelligence and modern machine learning techniques on the study of complex diseases in light of the vast amount of human health data now available. With over 200 million electronic medical records and 10 million genotyped individuals, humans have become the most extensively characterized species on Earth. The roundtable will explore how the integration of AI and dynamic systems theory can facilitate the transition from cross-sectional biobanks to truly longitudinal studies. Participants will discuss the potential of real-world, longitudinal data to infer causality, and relevant time scales, bridging the gap between molecular biology and our understanding of complex diseases. The conversation will also examine the prospects for identifying novel therapeutic targets with disease-modifying potential, exploring innovative therapeutic modalities, and utilizing generative chemistry to create end-to-end pipelines leading to groundbreaking drugs that revolutionize healthcare and disease prevention and treatments.
Ph.D. from the University of Amsterdam. Co-founder of Gero, a data-driven longevity biotech company that develops new drugs against aging and other complex diseases using AI-platform. An author of 75+ published papers in multiple domain areas, including publications in Science and Nature Communications.
Poster Session by Elucidata
Building Solid Data Foundations for AI/ML with Large Language Models (LLMs)
- To effectively train predictive models in drug discovery, large volumes of clean and linked data are required, which can be a costly and time-consuming task to curate manually. As such, there is a growing need for automated curation processes that can accurately and efficiently label data at scale.
- Elucidata has developed a biocuration process that leverages domain-trained BERT-like models for a variety of information extraction tasks: Identification of cell type, cell line, tissue, disease, and other characteristics from unstructured biomedical datasets. This approach has shown promising results in improving the quality and efficiency of data curation.
- In this session, we will explore the specifics of Elucidata’s NLP-based biocuration and how it can help R&D teams make their data interoperable, enable data and metadata integration and generate model quality datasets for AI/ML use cases.
- As a bonus, we will also demonstrate our experiments with Open AI’s Chat-GPT and its potential in solving edge cases in biocuration.
Dr. Abhishek Jha was an early member of the platform team at Agios Pharmaceuticals and supported multiple drug discovery programs, two of which have been approved by the FDA. As a founder of Elucidata, he is committed to building a transformative biotech company for the future that will provide clean and linked machine learning-ready data at every stage of drug discovery.
Evening Drinks Reception