ATS Seminar Series: Erik Schultes

The Advanced Technologies Section (ATS) of the National Center for Computational Sciences at ORNL is a world leader in developing and deploying scientific and technical solutions for leadership-class computing environments. The R&D activities of ATS are organized around designing and deploying leadership class systems, developing artificial intelligence solutions for science and smart facilities of the future, and stewardship of data and workflows at scale to enable science. The ATS Seminar Series is a forum for learning from experts and engaging with collaborators to advance their scientific mission.

Seminar Title:

Getting Ahead of the Virus with Machine Learning

The COVID-19 pandemic has entered a new phase: the emergence and spread of variants. ‘Variants of Concern’ are those that either escape the immune response induced by earlier infection or by targeted vaccination. Rather than passively waiting until such variants to occur in the population, we have proposed a data-intensive approach that combines computer simulation with real-world detection as input to machine learning, with the objective of predicting variants of concern before they are actually observed in populations. This grand challenge, if met, could lead pro-active public-health interventions and smarter pandemic response.

The approach is predicated on the computational and semantic interoperation of two kinds of data: (1) large-scale computer simulations predicting the structural and functional implications of large numbers of possible (but not yet observed) variants in the SARS-CoV-2 Spike protein and (2) low-cost, high-throughput mass spectroscopy detection of Spike protein variants that opens the door to high spacial and temporal resolution in the surveillance of Spike protein variants in the population. To ensure the automated reusability of these data (and these data with other large sequence and immunological datasets) we adhere closely to the FAIR Guiding Principles, treating each variant (whether theoretical or observed) as a FAIR digital twin. We also advocate for a pre-competitive, FAIR Immunomics Platform where variant digital twins are accessible to the research community at large, supporting both independent and collaborative development of machine learning approaches to variant prediction.

In this presentation, Dr. Schultes will introduce the problem use case, theoretical and computational considerations around large scale protein structure prediction, and an innovative mass spectroscopy approach to the detection of protein variants.

Speaker:

Erik Schultes

FAIR Implementation Lead, GO FAIR Foundation

Since July 2020, Dr. Schultes is the FAIR Implementation Lead at the GO FAIR Foundation.  Before assuming that position, he was co-founder of, and served as the International Science Coordinator for the GO FAIR International Support and Coordination Office (2017-2020). In this role, he worked with a diverse community of stakeholders to kick-start the development of FAIR data and services. Dr. Schultes is the architect of the Three-Point FAIRification Framework that is now used throughout Europe, Africa and the United States.  Dr. Schultes is also a member of the Leiden Center for Data Science at Leiden University and the Leiden Academic Center for Drug Discovery.  He is trained as an molecular evolutionary biologist with ongoing interests in data-intensive research questions.

Recorded Presentation:

Vimeo link: https://vimeo.com/579883368