About

Alberto Bartesaghi, PhD, is an Associate Professor of Computer Science and Biochemistry at Duke University, a position he has held since 2018. He earned his PhD in Electrical and Computer Engineering from the University of Minnesota in 2005. Following his doctoral studies, he served as an Associate Scientist at the National Cancer Institute at the National Institutes of Health in Bethesda, MD from 2005 to 2018. His research lab focuses on developing computational methods to solve the structure of large macromolecular complexes using advanced imaging techniques such as single particle cryo-electron microscopy, cryo-electron tomography, and sub-volume averaging. Beyond structural biology, his interests extend to machine learning, computer vision, image processing, and high-performance computing, reflecting a broad expertise in computational approaches to biological and biomedical challenges.

Research topics

• Biology
• Artificial Intelligence
• Computational biology
• Virology
• Computer Science
• Computer vision
• Physics
• Mathematics
• Optics
• Molecular biology
• Biochemistry
• Microbiology
• Genetics
• Cell biology
• Immunology

Selected publications

• nextpyp/prismpyp: v1.0.0

  Zenodo (CERN European Organization for Nuclear Research) · 2026-02-10

  otherOpen accessSenior author

  Power-spectrum and image domain learning for self-supervised micrograph evaluation

  PublisherDOI

• At FullTilt: Real-Time Open-Set 3D Macromolecule Detection Directly from Tilted 2D Projections

  arXiv (Cornell University) · 2026-04-12

  preprintOpen accessSenior author

  Open-set 3D macromolecule detection in cryogenic electron tomography eliminates the need for target-specific model retraining. However, strict VRAM constraints prohibit processing an entire 3D tomogram, forcing current methods to rely on slow sliding-window inference over extracted subvolumes. To overcome this, we propose FullTilt, an end-to-end framework that redefines 3D detection by operating directly on aligned 2D tilt-series. Because a tilt-series contains significantly fewer images than slices in a reconstructed tomogram, FullTilt eliminates redundant volumetric computation, accelerating inference by orders of magnitude. To process the entire tilt-series simultaneously, we introduce a tilt-series encoder to efficiently fuse cross-view information. We further propose a multiclass visual prompt encoder for flexible prompting, a tilt-aware query initializer to effectively anchor 3D queries, and an auxiliary geometric primitives module to enhance the model's understanding of multi-view geometry while improving robustness to adverse imaging artifacts. Extensive evaluations on three real-world datasets demonstrate that FullTilt achieves state-of-the-art zero-shot performance while drastically reducing runtime and VRAM requirements, paving the way for rapid, large-scale visual proteomics analysis. All code and data will be publicly available upon publication.

  PublisherDOI

• At FullTilt: Real-Time Open-Set 3D Macromolecule Detection Directly from Tilted 2D Projections

  arXiv (Cornell University) · 2026-04-12

  articleOpen accessSenior author

  Open-set 3D macromolecule detection in cryogenic electron tomography eliminates the need for target-specific model retraining. However, strict VRAM constraints prohibit processing an entire 3D tomogram, forcing current methods to rely on slow sliding-window inference over extracted subvolumes. To overcome this, we propose FullTilt, an end-to-end framework that redefines 3D detection by operating directly on aligned 2D tilt-series. Because a tilt-series contains significantly fewer images than slices in a reconstructed tomogram, FullTilt eliminates redundant volumetric computation, accelerating inference by orders of magnitude. To process the entire tilt-series simultaneously, we introduce a tilt-series encoder to efficiently fuse cross-view information. We further propose a multiclass visual prompt encoder for flexible prompting, a tilt-aware query initializer to effectively anchor 3D queries, and an auxiliary geometric primitives module to enhance the model's understanding of multi-view geometry while improving robustness to adverse imaging artifacts. Extensive evaluations on three real-world datasets demonstrate that FullTilt achieves state-of-the-art zero-shot performance while drastically reducing runtime and VRAM requirements, paving the way for rapid, large-scale visual proteomics analysis. All code and data will be publicly available upon publication.

  PublisherOA PDF

• prismPYP: Power-spectrum and image domain learning for self-supervised micrograph evaluation

  Structure · 2026-03-01

  articleOpen accessSenior author

  High-throughput data collection in single-particle cryo-electron microscopy (EM) necessitates fast, accurate, and generalizable methods to assess micrograph quality. Manual micrograph curation scales poorly to large datasets and often misclassifies images due to sample-specific variability. Fully supervised deep-learning methods show promise in scalability and feature learning. However, dependence on annotated data limits generalizability. We present prismPYP, a self-supervised, data-driven framework that uses domain-specific image augmentations to perform label-free feature learning on micrographs and power spectra. From the learned, low-dimensional image representations, we perform feature-based image clustering that reveals distinct and consistent indicators of image quality. For validation, we used the resulting high-quality images to determine high-resolution structures that matched the quality of maps determined using manual curation, but using fewer particles. prismPYP generalizes across experimental conditions, imaging hardware, and both conventional single-particle and time-resolved cryo-EM. It is both interpretable and computationally efficient, and enables rapid, scalable quality assessment for cryo-EM micrographs.

  PublisherDOI

• nextpyp/prismpyp: v1.0.0

  Zenodo (CERN European Organization for Nuclear Research) · 2026-02-10

  otherOpen accessSenior author

  Power-spectrum and image domain learning for self-supervised micrograph evaluation

  PublisherDOI

• Image-processing methods for electron microscopy of biological specimens

  Acta Crystallographica Section D Structural Biology · 2025-05-19

  editorialOpen access

  The focused issue on Image-processing methods for electron microscopy of biological specimens is introduced. The virtual issue is available at https://journals.iucr.org/special_issues/2025/imageprocessing.

  PublisherOA PDFDOI

• Tools to Streamline <i>in situ</i> Structural Analysis Using Cryo-ET

  Microscopy and Microanalysis · 2025-07-01

  articleOpen access1st authorCorresponding
  PublisherOA PDFDOI

• Strategies for studying discrete heterogeneity in situ using cryo-electron tomography

  Current Opinion in Structural Biology · 2025-11-15

  articleOpen access1st authorCorresponding

  Structural variability plays a crucial role in enabling biological function, as the ability of proteins to adopt multiple conformations allows them to perform diverse cellular tasks. Cryo-electron tomography combined with subtomogram averaging and classification has emerged as a powerful technique for elucidating the conformational dynamics of proteins in their near-native environment. Increased data availability has provided a driving force for improvements in image classification algorithms which have enabled conformational heterogeneity studies of proteins in situ at higher resolutions than previously possible. In particular, the use of 2D particle projections extracted from raw tilt-series paired with constrained classification strategies of projection sets has emerged as a promising strategy for classifying particles in 3D. Despite these efforts, further method development will be needed to extend the applicability of current strategies for 3D classification to more challenging biological targets, including low-molecular weight complexes and membrane proteins. • Computational tools for heterogeneity analysis allow studying protein conformational variability in the cellular environment. • 3D classification of subtomograms represents a versatile tool for probing protein conformational variability landscapes. • Constrained classification of 2D projection sets is a promising strategy for studying discrete and continuous variability. • Particle alignment combined with discrete classification enables high-resolution heterogeneity analysis in situ.

  PublisherDOI

• Data-Driven Evaluation of Cryo-EM Micrographs Using Self-Supervised Learning

  Microscopy and Microanalysis · 2025-07-01

  articleSenior author
  PublisherDOI

• Raw tilt series of HERV-K Gag immature lattice

  EMPIAR dataset · 2025-02-28

  datasetOpen accessSenior author

  EMPIAR, the Electron Microscopy Public Image Archive centered at EMBL-EBI, is a public resource for raw electron microscopy images related to EMDB, contains micrographs, particle sets and tilt-series.

  PublisherDOI

Frequent coauthors

• Sriram Subramaniam

  201 shared

• Mario J. Borgnia

  National Institutes of Health

  162 shared

• Alan Merk

  Frederick National Laboratory for Cancer Research

  68 shared

• Jacqueline L.S. Milne

  National Cancer Institute

  67 shared

• Doreen Matthies

  50 shared

• Prashant Rao

  Beth Israel Deaconess Medical Center

  50 shared

• Joel R. Meyerson

  Weill Cornell Medicine

  42 shared

• Soojay Banerjee

  National Cancer Institute

  40 shared

Labs

• Bartesaghi LabPI