Pooja Sharma
Principal Scientist Amgen Inc.
Pooja Sharma, Principal Scientist at Amgen, is an accomplished discovery scientist with extensive expertise in biophysical, biochemical, and structural techniques spanning small molecules, peptides, membrane proteins, and RNA targets. She brings deep experience in SPR, NMR, X-ray crystallography, DSF/TSA, ITC, MST, and AI-enabled analyses, alongside broad proficiency in soluble and membrane protein expression systems, including SMALPs and nanodiscs. Pooja has contributed to fragment-based lead discovery at Harvard Medical School and the Walter and Eliza Hall Institute and has enabled multiple early discovery programs by overcoming complex mechanistic and assay challenges. Her current work focuses on advancing alternative biophysical approaches for challenging membrane targets.
Seminars
Artificial intelligence and machine learning are transforming the way biophysical data is generated, interpreted, and applied. But beyond the day-to-day integration of AI into experimental workflows, what are the broader opportunities, challenges, and limitations?
This interactive panel brings together leading experimentalists, computational biophysicists, and AI specialists to explore the bigger picture and provides you the opportunity to have honest conversations with your industry colleagues about hype vs. value.
Key Discussion Themes:
- Where can AI add true value in biophysics beyond data processing? Predictive modelling, hit triaging, protein design?
- Which areas of drug discovery are most ready for AI adoption and where should caution succeed?
- The challenges of AI adoption including data curation, standardization, reproducibility, interpretability, and integration with lab workflows
- The evolving role of experimental validation alongside AI predictions, how to ensure we maintain scientific rigor
Blue Sky Thinking & Audience Feedback: In 5 years, what do you hope AI/ML has enabled / transformed the way you do something?
As biophysics workflows generate increasingly large and complex datasets, AI and machine learning are emerging as transformative tools to enhance signal extraction, automate analysis, and predict experimental outcomes. But how can we integrate these tools effectively, and responsibly, into the experimental pipeline?
This interactive workshop will bring together experts across experimental biophysics, computational modelling, and data science to explore the practical realities, opportunities, and pitfalls of implementing AI in biophysical research and drug discovery.
Join your peers to collaboratively examine:
- How AI and machine learning can streamline data processing, noise reduction, and artefact identification in high-throughput biophysical screening
- Opportunities for predictive modelling to anticipate aggregation, stability, or binding profiles
- What it means to make biophysical data AI-ready: standardization, metadata, and dataset curation
- How to validate AI-driven insights and maintain scientific rigor
- Lessons learned from early adopters applying AI to techniques like SPR, BLI, and NMR
Why Take Part?
Walk away with a practical understanding of where AI integration can genuinely accelerate discovery, and where caution, curation, and collaboration are still required.
- Evaluate emerging label-free and solution-based techniques suited for membrane proteins
- Discuss how new detection formats improve signal quality and stability for hydrophobic systems
- Learn how to select orthogonal assays that overcome traditional immobilization limitations
Biophysics offers an abundance of techniques and assays, each with their own benefits and challenges, but knowing which to use, when to use it, and how to evolve your assay strategy across discovery phases is a constant challenge. This interactive workshop brings together early discovery scientists from pharma and biotech to share practical decision-making frameworks for building high-confidence assay cascades.
Join this workshop to:
- Explore practical strategies for designing cascades that evolve with program maturity
- Uncover exclusive insights into trade-offs between throughput, resolution, and confidence in hit identification
- Discuss approaches for choosing between overlapping or orthogonal assays (e.g., SPR vs BLI vs ITC)
- Hear real-world examples from both pharma and biotechs of how assay source (fragment screens, DEL, HTS) and target type influence cascade design
- Extract tips to interpret conflicting data and avoid common pitfalls
Why Take Part?
Leave with practical insights on how to design, adapt, and troubleshoot their own screening workflows according to target class, modality, throughput needs, and data quality.
