Since the inception of biomedical informatics, it has been recognized that the vast amount of data generated by large-scale genomics technologies presents a significant opportunity for artificial intelligence and data science (i.e, informatics). This has been evident with the widespread phenomena of OpenAI's ChatGPT and becomes even more apparent when additional data such as electronic health records, pathology reports, clinical data, drug response data, and annotation data are included.
Many academic and commercial organizations around the world have been trying to glean insights and conclusions from large genomic data sets through the use of genome-wide association studies (GWAS) and sequencing technologies. To have confidence in the data and integrate it into clinical practice on a routine basis, we need to improve the diversity of genomic data sets to eliminate any hidden biases, standardize the tools used to monitor the performance of AI-based systems, streamline the integration with patient care solutions, and develop computational methods that enable the interpretation of genomic data at the individual patient level without overwhelming physicians with large and high-dimensional data.
"I believe that we have to depend on advances in data sciences in order to expand on what is for the moment a fairly limited repertoire of computational solutions for things like polygenic risk scores for diseases definition."
Russ Cucina, Chief Health Information Officer & VP Genetic and Genomic Services, UCSF Health System + PMWC January 25-27, 2023 AI & Data Sciences in Clinical Practice Track Chair
While there are a few polygenic scores available for cardiac coronary artery disease, hypertension, and a few other conditions, we currently only have a limited number of computational approaches that convert an individual's genomic information into actionable risk scores and corresponding interventions. Although AI techniques are already being used in some clinical applications, their use in genetics and genomics is still in its infancy.
There is great potential for AI and data science in genomics/genetics applications, and for this reason, we have created a track that focuses on these critical aspects, including: ▲The potential benefits and risks of preemptive testing and who should pay for it at the individual and population level, ▲ How to establish genomics in the clinic as a routine through the use of AI and automation, ▲ How to best translate omics data into clinical practice via interoperable genomic laboratory systems, and ▲ How to improve equity in genomic data
The beauty of PMWC is its diverse audience, which is necessary to address the various challenges faced by AI and data science in genomics and genetics. For this reason, we have invited leaders from various industry sectors to actively contribute to this important topic, including medical researchers, physicians, genetic counselors, genomic laboratory system providers, laboratory representatives, members of commercial testing labs, big data analytics providers, and healthcare providers.
AI & Data Science in Clinical Practice track highlights - January 25 9.00 AM - 4.00 PM:
How can AI and Data Science Determine Which Patients should be Preemptively Tested? – a panel chaired by Russ Cucina, UCSF with Lisa Kroon (UCSF), Eran Segal (Weizmann Institute), and Daniel Knecht (CVS Health)
Establishing Genomics in the Clinic as a Routine Requires AI and Automation – chaired by Bani Tamraz (UCSF) with Sergio Baranzini (UCSF), Gill Bejeran (Stanford University), and James Shalaby (Elimu Informatics)
Representation and Equity in Genomic Data – chaired by Karen Miga (UCSC). This panel includes Timothy Rebbeck (Harvard University), Malcolm John (UCSF), and Keolu Fox (UCSD)
Overcoming Barriers to Bringing Multi-Omic's into Clinical Practices – a panel chaired by Russ Cucina (UCSF) with Peter DeVault (Epic), Kevin Hass (Myriad), and Jim Chen (Tempus)
Misaligned Incentives Across Public-Private Partnerships in Patient Health – a panel chaired by Justin Wildsmith, Syntropy) with Ken Harris (AWS), Steve Goldstein (UC Irvine), Suzanne George (Dana Farber Cancer Institute), and Nelson D’Antonio (Palantir)
PMWC AI & Data Science Showcases:
A*STAR - I²R, Aiforia, AllStripes, Amazon Web Services (AWS), AMPEL BioSolutions, BC Platforms, DrugBank, Duality Technologies, Flywheel, Genialis, Healx, Katana Graph, Medidata, Nucleai, Omics Data Automation, OneThree Biotech, Pachyderm, Pangea Biomed, Rosalind, Seven Bridges, Tag.bio, Tempus, Valley Children's Healthcare, and the Weizmann Institute of Science
