Scientist - Structural Biology Machine Learning

Astera is a private foundation on a mission to steer science and technology toward an abundant future. We believe the coming years will bring an era of unprecedented scientific and technological advancement as exponential progress in AI converges with central advances in other fields to dramatically accelerate innovation. This inflection point provides an unparalleled opportunity to fundamentally rethink the institutions, systems, and tools that drive scientific progress.

Unlike traditional non-profit research organizations, projects supported by Astera operate like high-velocity startups, allowing us to focus on ambitious goals, match structure to problem, and attract strong technical talent and leadership. You can read more about our mission, vision, and programming here.

The diffUSE Project is looking for a scientist to build machine learning algorithms to extract protein conformational dynamics directly from experimental structural biology data (i.e. electron density, structure factors, diffraction patterns, etc.). The diffUSE Project is an ambitious initiative designed to unlock our understanding of protein dynamics at scale by strengthening experimental methods, computational models, and global infrastructure. Our goal is to establish dynamic structural biology as a foundational pillar of modern science, as transformative and indispensable as static structures has been for biology.

This role sits at the intersection of generative modeling and experimental structural biology. You would develop algorithms that treat experimental observables (i.e. electron density, structure factors, diffraction patterns, etc.) as direct training inputs to improve the fitting of dynamic structural biology models, design evaluation frameworks to assess them, and work closely with structural biologists and biophysicists who generate the data. We are looking for someone who is comfortable with physics-constrained learning or inverse problems, and looking to join an ambitious team to improve our models and understanding of dynamics.

This is a full-time position within the diffUSE Project, in-person at Radial, a division of the Astera Institute.

• →

  Develop and own novel algorithmic approaches to extracting protein dynamics from experimental observables

• →

  Experience with probabilistic/Bayesian methods

• →

  Inverse problems experience

• →

  Design, train, and deploy open-source ML models that learn directly from experimental X-ray crystallography data (structure factors, electron density, diffraction patterns) for conformational ensemble modeling

• →

  Develop and benchmark metrics for conformational ensemble modeling and comparison against experimental data

• →

  Collaborate with domain scientists to integrate outputs with experimental pipelines and refine hypotheses in an iterative design–test–learn loop

• PhD in Data Science, Computer Science, Bioinformatics, Biophysics, Computational Chemistry, or related field

• Deep experience with generative models (diffusion, flow matching), probabilistic inference and/or representation learning (e.g., contrastive learning)

• Ability to work effectively in a multidisciplinary team environment