Deirdre Costello
Associate Director, Structural Biology Genesis Molecular AI
Deirdre Costello is an accomplished professional in structural biology, currently serving as the Associate Director at Genesis Therapeutics since March 2025. Prior experience includes the role of Principal Scientist at Treeline Biosciences and multiple positions at Eli Lilly and Company, where Deirdre advanced from Research Scientist to Senior Research Scientist. Academic credentials include a PhD and MS in Chemical and Biomolecular Engineering from Cornell University, complemented by a BS in Chemical Engineering from University College Dublin. Deirdre’s doctoral research focused on the fusion kinetics of influenza, contributing to the development of anti-fusogenic therapeutics. Additional experience includes a postdoctoral position at Harvard Medical School and instructional roles fostering engineering education for youth.
Seminars
- Comparing access, cost, and timelines for large pharma versus biotechs, and when outsourcing or academic partnerships make more sense
- Addressing challenges around data volume, selection, and cloud storage
- How data processing, classification, and refinement workflows impact turnaround time and confidence in results
- Current constraints including size limits, flexibility, heterogeneity, and challenges with smaller molecules or weak interactions
- Emerging hardware, sample preparation strategies, and computational advances aimed at improving resolution, throughput, and applicability
Cryo-EM is now a routine part of many discovery pipelines, yet misunderstanding around how data is generated, what its limitations are, and how long it takes, continue to create friction between structural biologists and the wider discovery team.
This workshop is designed to give non-specialists and structural scientists working cross-functionally; a clear, realistic understanding of the Cryo-EM workflow, from sample preparation and data collection through to model interpretation, without turning the session into a technical training course. The goal is not to teach attendees how to run a microscope or process raw data, but to help them understand what is happening behind
the scenes, why certain experiments succeed or fail, and how that impacts timelines
and decision confidence.
By building shared literacy across structural biology, computational chemistry, medicinal chemistry, and biology, this workshop aims to improve feedback cycles, reduce unnecessary structural effort, and help teams ask better questions of Cryo-EM, before, during, and after data generation.
Workshop Highlights:
- A high-level walkthrough of sample preparation, data collection, and reconstruction, focused on why these steps matter for timelines, resolution, and interpretability
- Understanding the limits of Cryo-EM data and how this can affect downstream design decisions
- Exploring how the availability of predictive models changes when experimental structures are needed, and when cryo-EM is essential to resolving uncertainty
- Helping non-Cryo specialists understand what they can (and cannot) conclude from Cryo-EM data to improve overall communication between structural biologists, chemists and modelers
