[staff-physics-and-astronomy-HL-UHD-UD] Talk by Alisia Fadini (University of Cambridge) – Monday, May 19: ROCKET – AlphaFold Augmentation for Experimental Data Integration

[Beta NAT Secretariaat]

Dear all,

On Monday, May 19th, we are pleased to welcome Alisia Fadini from University of Cambridge, who will be visiting the VU and giving a seminar primarily aimed at structural biology students.

Date: Monday, May 19
Time: 13:30–15:15
Location: NU-Theatre 9

Alisia is currently working with the groups of Randy Read (University of Cambridge) and Mohammed AlQuraishi (Columbia University) at the cutting edge of machine learning–driven protein structure prediction.

Title:  ROCKET: an AlphaFold augmentation integrating crystallographic & cryoEM/ET data
Structural biology is in an era of dynamics and macromolecular assemblies, but turning raw experimental data into atomic models at scale remains challenging. We introduce ROCKET: an AlphaFold augmentation that integrates crystallographic and cryoEM/ET data, with room for more! AF-based methods encode rich structural priors but lack a general mechanism for integrating arbitrary data modalities. ROCKET tackles this by optimizing latent representations to fit experimental data at inference time, without retraining. ROCKET integrates experimental likelihood targets within AlphaFold’s differentiable prediction pipeline to optimize MSA profile features. Structure refinement becomes a search in evolutionary space instead of Cartesian space. What does this unlock? ROCKET samples barrier-crossing conformations that standard refinement methods often fail to reach, such as ligand-induced loop rearrangements or domain shifts.
The low-resolution challenge is also particularly key for emerging cryo-ET data. We find that ROCKET remains robust in this regime and extracts different conformations from sub-tomogram averages.
ROCKET performs a new type of structure refinement by optimizing latent representations in evolutionary space. This unlocks possibilities for high-throughput ligand screening, assemblies solved at low resolution, and conformational landscapes – bringing automation towards new frontiers.

After the seminar, there will be a borrel/drinks, offering a chance to chat with the speaker in a more informal setting. Everyone is welcome!



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://listserver.vu.nl/pipermail/senior-staff-physics-and-astronomy/attachments/20250514/8e68ec94/attachment.html>